WO2010109809A1 - Failure prediction system, electronic device, and failure prediction method - Google Patents

Failure prediction system, electronic device, and failure prediction method Download PDF

Info

Publication number
WO2010109809A1
WO2010109809A1 PCT/JP2010/001865 JP2010001865W WO2010109809A1 WO 2010109809 A1 WO2010109809 A1 WO 2010109809A1 JP 2010001865 W JP2010001865 W JP 2010001865W WO 2010109809 A1 WO2010109809 A1 WO 2010109809A1
Authority
WO
WIPO (PCT)
Prior art keywords
failure prediction
value
amplitude value
prediction system
failure
Prior art date
Application number
PCT/JP2010/001865
Other languages
French (fr)
Japanese (ja)
Inventor
葛西茂
佐々木康弘
酒井浩
Original Assignee
日本電気株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 日本電気株式会社 filed Critical 日本電気株式会社
Priority to JP2011505853A priority Critical patent/JP5573834B2/en
Publication of WO2010109809A1 publication Critical patent/WO2010109809A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G11—INFORMATION STORAGE
    • G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
    • G11B5/00—Recording by magnetisation or demagnetisation of a record carrier; Reproducing by magnetic means; Record carriers therefor
    • G11B5/40—Protective measures on heads, e.g. against excessive temperature 
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
    • G01H1/00—Measuring characteristics of vibrations in solids by using direct conduction to the detector
    • G—PHYSICS
    • G11—INFORMATION STORAGE
    • G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
    • G11B19/00—Driving, starting, stopping record carriers not specifically of filamentary or web form, or of supports therefor; Control thereof; Control of operating function ; Driving both disc and head
    • G11B19/02—Control of operating function, e.g. switching from recording to reproducing
    • G11B19/04—Arrangements for preventing, inhibiting, or warning against double recording on the same blank or against other recording or reproducing malfunctions
    • G11B19/041—Detection or prevention of read or write errors
    • G11B19/042—Detection or prevention of read or write errors due to external shock or vibration
    • G—PHYSICS
    • G11—INFORMATION STORAGE
    • G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
    • G11B19/00—Driving, starting, stopping record carriers not specifically of filamentary or web form, or of supports therefor; Control thereof; Control of operating function ; Driving both disc and head
    • G11B19/02—Control of operating function, e.g. switching from recording to reproducing
    • G11B19/04—Arrangements for preventing, inhibiting, or warning against double recording on the same blank or against other recording or reproducing malfunctions
    • G11B19/048—Testing of disk drives, e.g. to detect defects or prevent sudden failure

Definitions

  • the present invention relates to a failure prediction system, and more particularly to a failure prediction system for electronic devices and the like.
  • Patent Document 1 describes a failure prediction system in a magnetic disk device.
  • FIG. 19 is a block diagram showing the configuration of the magnetic disk device.
  • the magnetic disk apparatus 101 includes a read channel 106 having automatic amplification adjustment (AGC), a memory 109, a disk medium 115, a head monitoring unit 117, and a failure prediction unit 118. At least.
  • AGC automatic amplification adjustment
  • AGC refers to an adjustment mechanism that automatically amplifies a signal output from the head when servo data is read from the disk medium 115. For example, when the head is lifted and separated from the disk medium 115, the magnetic field generated at the head tip is weakened on the disk medium 115, so that the output signal from the weakened head is automatically amplified.
  • the failure prediction unit 118 detects the abnormal flying of the head based on whether or not the AGC signal output intensity is within a predetermined range, that is, whether or not the gain is in a state where reading and writing can be normally performed.
  • the head monitoring unit 117 monitors the value of the AGC provided in the read channel 116 for a part of measurement zones provided in the disk medium 115 when a certain timing for measuring the number of abnormal flying of the head comes. taking measurement.
  • the failure prediction unit 118 detects that the head is abnormally flying. Then, the failure prediction unit 118 stores in the memory 109 the cumulative number of abnormal levitations of the head and the number of abnormal levitations for each measurement time period (for example, classifications A, B, C, and D from the oldest at the time of measurement). save. The failure prediction unit 118 creates a graph including the measurement time segment from the time of shipment and the number of abnormal ascents from the measurement data stored in the memory 9. Then, the failure prediction unit 118 calculates whether the abnormal ascent number tends to increase or the accumulated number of abnormal ascent times from the shipment does not exceed the threshold value.
  • the failure prediction unit 118 confirms whether or not the abnormal flying frequency of the head at the previous measurement recorded in the memory 109 is increasing when the abnormal flying frequency is increasing.
  • the failure prediction unit 118 issues a warning that there is a high possibility that a failure will occur if the number of abnormal flying of the head tends to increase even during the previous measurement.
  • the failure prediction unit 118 checks whether or not the cumulative number of abnormal levitation times recorded in the memory 109 exceeds the threshold when the abnormal levitation frequency does not tend to increase. If the cumulative number of abnormal ascents exceeds the threshold, the failure predicting unit 118 issues a warning that there is a high possibility that a failure will occur.
  • Patent Document 2 describes an abnormality detection device that diagnoses an abnormality by detecting AE (Acoustic Emission) from a bearing.
  • AE Acoustic Emission
  • AE is a phenomenon in which strain energy stored in the material is released as an elastic wave when the material is deformed or a crack is generated.
  • the abnormality detection apparatus includes a peak hold circuit 206, a reference value generator 207, a comparator 208, an AD converter 210, and a CPU (central processing unit) 211.
  • the peak hold circuit 206 detects the AE signal
  • the peak hold circuit 206 detects and holds the peak value of the AE signal, and outputs the peak value to the AD converter 210 and the comparator 208.
  • the comparator 208 compares the reference value generated by the reference value generator 207 with the peak value from the peak hold circuit 206. When the peak value exceeds the reference value, the trigger signal is sent to the AD converter 210 and the CPU 211. Output to.
  • the AD converter 210 is ready to process the AE signal when it receives the trigger signal from the comparator 208, and AD converts the AE signal.
  • the CPU 211 is also in a state where the AE signal can be processed when receiving the trigger signal from the comparator 8, and takes the digitized AE signal from the AD converter 210 and diagnoses a bearing abnormality. Therefore, the AD converter 210 and the CPU 211 operate to process the AE signal only when the peak value of the AE signal exceeds the reference value.
  • the AD converter 210 and the CPU 211 do not perform an operation to process the AE signal when the peak value of the AE signal is lower than the reference value. That is, since the AD converter 210 and the CPU 211 do not perform an operation for signal processing of the AE signal when the AE signal is a value that is not effective for bearing abnormality diagnosis, the normal operation performance is not deteriorated and the failure prediction accuracy is high.
  • a failure prediction system can be provided.
  • An object of the present invention is to provide a failure prediction system that suppresses a decrease in the operation performance of an apparatus without increasing the frequency of failure prediction.
  • the failure prediction system includes a vibration measuring instrument that measures vibration generated from a device, and a failure prediction when an amplitude value of the vibration measured by the vibration measuring instrument exceeds a predetermined threshold and continues for a predetermined time or more.
  • a signal processor to perform.
  • the present invention can suppress a decrease in the operating performance of the apparatus without increasing the frequency of failure prediction.
  • FIG. 10 is a configuration diagram using an external memory of an external device when a failure prediction system according to a fifth embodiment of the present invention is built in a computer device.
  • a failure prediction system according to a sixth embodiment of the present invention is externally attached to a computer device, it is a configuration diagram using a device memory of the computer device.
  • the failure prediction system according to the sixth embodiment of the present invention is externally attached to a computer device, it is a configuration diagram using an external memory of an external device.
  • the failure prediction system according to the seventh embodiment of the present invention is built in a hard disk device, it is a configuration diagram using a device memory of the hard disk device. It is a failure prediction result of the memory
  • storage device based on an Example. 1 is a configuration diagram of a magnetic disk device in Patent Document 1.
  • FIG. 10 is a configuration diagram of a bearing abnormality detection device in Patent Document 2.
  • FIG. 1 is a configuration diagram of a magnetic disk device in Patent Document 1.
  • FIG. 1 shows a configuration diagram of the failure prediction system 1 of the present invention.
  • the failure prediction system 1 of the present invention includes a vibration measuring instrument 3, a signal processor 7, and a memory 8.
  • the vibration measuring instrument 3 has a function of measuring the vibration of a device that is a target of failure prediction, and is attached to a device that is not shown.
  • a device refers to an electronic device equipped with a device that generates vibration, such as a hard disk device.
  • the signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6.
  • the signal conversion unit 4 is connected to the vibration measuring instrument 3 and the signal analysis unit 5 and has a function of converting an analog signal of vibration into a digital signal.
  • the signal analysis unit 5 is connected to the signal conversion unit 4 and measures the amplitude value of the vibration digital signal based on the vibration spectrum viewed from the frequency value. As shown in FIG. 3, the vibration spectrum is a graph with the vertical axis representing the vibration amplitude value and the horizontal axis representing the frequency. When the vibration amplitude value exceeds a predetermined threshold at a specific frequency f and continues for a predetermined time or longer, the signal analysis unit 5 analyzes the time change of the amplitude value at the specific frequency f.
  • the signal determination unit 6 is connected to the signal analysis unit 5 and has a function of performing failure prediction based on a temporal change in the amplitude value of vibration at a specific frequency f.
  • the memory 8 is connected to the signal analysis unit 5.
  • the vibration measuring device 3 measures the frequency and vibration amplitude value of the device, and sends a signal of the measured frequency and vibration amplitude value to the signal conversion unit 4 of the signal processor 7.
  • S2 processing is advanced.
  • the signal conversion unit 4 converts an analog signal of vibration from the vibration measuring instrument 3 into a digital signal and transmits the digital signal to the signal analysis unit 5. Next, the process proceeds to S3.
  • the signal analysis unit 5 analyzes the vibration digital signal from the signal conversion unit 4 using a vibration spectrum in which the magnitude of the amplitude value is viewed from the frequency value. Next, the process proceeds to S4.
  • the signal analysis unit 5 stores data in which the amplitude value of vibration exceeds a predetermined threshold A in the memory 8.
  • the signal analysis unit 5 advances the process to S7. If the vibration amplitude value exceeds the threshold A and does not continue for more than the threshold time B, the process returns to S1. In addition, when returning to the process of S1, the content preserve
  • the threshold value A of the amplitude value in the analysis of the vibration spectrum starts the analysis of the failure prediction of the vibration not related to the failure so as to reduce the leakage with respect to the vibration that causes the failure, and the normal of the device. It is set so as not to affect the operation of.
  • the threshold time B that continues in a state in which the amplitude value exceeds the threshold value A, the leakage of the vibration that causes the failure is reduced, and the threshold value B also has vibrations that are not related to the failure. It is set so that failure prediction analysis is started and normal operation of the device is not affected.
  • the threshold value A of the amplitude value and the threshold time B of the duration time in which the threshold value A is exceeded are determined based on vibrations generated according to the failure factor of the device.
  • the amplitude value of the vibration spectrum when the device actually broke down and the duration of the amplitude value exceeding a certain threshold were measured from the experiment for each frequency and compiled into a database. Based on this database, a threshold A and a threshold time B that cause failure are set. In addition, since the cause of failure differs according to the frequency, the threshold A and the threshold time B are set according to the frequency.
  • the signal determination unit 6 starts failure prediction based on the temporal change of the vibration amplitude value at the specific frequency f.
  • the failure prediction system 1 in the first embodiment is such that the vibration analysis starting unit 9 is a value that the amplitude value of the vibration spectrum exceeds the threshold value A as shown in FIG. It monitors only whether it continues. That is, the failure prediction system 1 does not always perform failure prediction, but performs monitoring to directly extract characteristic vibrations accompanying the progress to failure as a first step.
  • the failure prediction system 1 can suppress the frequency of failure prediction by extracting the specific frequency f of vibration that is likely to cause a failure. Therefore, the failure prediction system 1 does not need to acquire and analyze vibration data of devices that are not involved in the progress of the failure, and can distinguish between sudden vibration and vibration accompanying the progress of the failure. As a result, the failure prediction system 1 can suppress a decrease in the performance of the target device by analyzing the failure prediction in spite of normality, thereby reducing the performance of arithmetic processing in the normal operation of the device. Therefore, the load can be suppressed.
  • Fig. 4 shows the overall system configuration.
  • the failure prediction system 1 includes a late signal measuring device 3, a signal processor 7, and a memory 8 as in the first embodiment.
  • the signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6.
  • the connection relationship between the blocks is the same as that in the first embodiment.
  • the signal analysis unit 5 includes a vibration analysis start unit 9 and a vibration amplitude value analysis unit 10.
  • the vibration analysis start unit 9 is connected to the signal conversion unit 4 and measures the amplitude value of the vibration digital signal from the vibration spectrum as seen from the frequency value. When the vibration amplitude value exceeds a predetermined threshold at a specific frequency f and continues for a predetermined time or longer, an analysis is started for a temporal change in the vibration amplitude value at the specific frequency f.
  • the vibration amplitude value analysis unit 10 is connected to the vibration analysis start unit 9, and performs linear regression with the time as an independent variable for the vibration amplitude value, and calculates the slope of the line every predetermined measurement time.
  • the vibration measuring device 3 measures the frequency and vibration amplitude value of the device, and sends a signal of the measured frequency and vibration amplitude value to the signal conversion unit 4 of the signal processor 7.
  • S102 processing is advanced.
  • the signal analysis start unit 9 in the signal analysis unit 5 performs the processes corresponding to S3 to S6 in S102. Then, when the signal analysis start unit 9 has exceeded the predetermined threshold A at the specific frequency f and the threshold time B, which is a predetermined duration, has elapsed, the process proceeds to S103.
  • the vibration amplitude value analysis unit 10 sets a lower limit value and an upper limit value of an amplitude value to be measured in order to exclude an abnormal point. Next, the process proceeds to S104.
  • the vibration amplitude value analysis unit 10 compares whether the measured amplitude value is within a range set between the lower limit value and the upper limit value. If the amplitude value exceeds the range, the process proceeds to S105. If the measured value does not exceed the range, the process proceeds to S106.
  • the vibration amplitude value analysis unit 10 excludes amplitude values exceeding the above range as abnormal data. Next, the process proceeds to S106.
  • the vibration amplitude value analysis unit 10 performs linear regression using the time as an independent variable for the amplitude value at a specific frequency f stored in the memory 8. Next, the process proceeds to S107.
  • the vibration amplitude value analysis unit 10 calculates the inclination for each predetermined number of consecutive points.
  • the process proceeds to S108.
  • a value that causes a failure is set on the basis of a database of experiments in which the device actually failed as in the case of the threshold A and the threshold time B.
  • the vibration amplitude value analysis unit 10 ends the failure analysis of the amplitude value at the specific frequency f when a predetermined time has elapsed, and returns to the process of S101. Note that when returning to the processing of S101, the content stored in the memory can be reset. If the predetermined time has not elapsed, the vibration amplitude value analysis unit 10 advances the process to S109.
  • the signal determination unit 6 sets a determination threshold value and a determination continuation number, and performs a failure determination from the transition of the slope of the linear regression. As shown in FIG. 6, the slope analysis by linear regression is calculated for every predetermined number of consecutive points.
  • the signal determination unit 6 observes the slope of each predetermined number of points calculated by the vibration amplitude value analysis unit 10 at a value that is equal to or greater than the determination threshold and exceeds the determination continuation count. The process proceeds to S110 and it is determined that a failure of the device can be predicted.
  • the determination threshold value is set to a specified number of times ⁇ times the slope a1 calculated at a predetermined number of points first calculated after the amplitude value at a specific frequency f is started.
  • the judgment continuation count is a value at which the slope of every predetermined number of consecutive measured data exceeds the judgment threshold value and is continuously generated.
  • the determination threshold and the determination continuation number are set to values that cause failure based on an experimental database in which the device actually failed.
  • the signal determination unit 6 sets a determination threshold value and a determination continuation number for vibrations involved in the progress of the failure, and determines failure from the transition of the slope of linear regression. By performing this, it is possible to perform failure prediction with high accuracy.
  • the vibration amplitude value analysis unit 10 can exclude vibrations caused by noise or the like by setting the upper and lower thresholds of the amplitude value to be measured in order to exclude abnormal points, and by a simple calculation process.
  • FIG. 7 shows an overall system configuration diagram.
  • the failure prediction system 1 includes a late signal measuring device 3, a signal processor 7, and a memory 8 as in the second embodiment.
  • the signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6, and the signal analysis unit 5 includes a vibration analysis start unit 9 and a vibration amplitude value analysis unit 10.
  • the connection relationship between the blocks is the same as that in the second embodiment.
  • the signal analysis unit 5 has a residual setting unit 11.
  • the residual setting unit 11 is connected to the vibration amplitude value analysis unit 10 and the signal determination unit 6, calculates a residual threshold value from a slope for each predetermined measurement time, and excludes residuals exceeding the residual threshold value. It has a function.
  • the vibration measuring device 3 measures the frequency and vibration amplitude value of the device, and sends a signal of the measured frequency and vibration amplitude value to the signal conversion unit 4 of the signal processor 7.
  • the process of S202 is advanced.
  • the signal conversion unit 4 and the vibration analysis start unit 9 perform a signal analysis start process similar to the operation in S102 of the second embodiment. Next, the process proceeds to S203.
  • the vibration amplitude value analysis unit 10 sets a lower limit value and an upper limit value of the amplitude value to be measured in order to exclude abnormal points. Next, the process proceeds to S204.
  • the vibration amplitude value analysis unit 10 compares whether the measured amplitude value is within a range set between the lower limit value and the upper limit value. If the amplitude value exceeds the range, the process proceeds to S205. If the measured value does not exceed the range, the process proceeds to S206.
  • the vibration amplitude value analysis unit 10 excludes measurement values exceeding the above range as abnormal data. Next, the process proceeds to S206.
  • the vibration amplitude value analysis unit 10 performs linear regression using the time as an independent variable for the amplitude value at a specific frequency f stored in the memory 8.
  • the process proceeds to S207.
  • the vibration amplitude value analysis unit 10 calculates the inclination for each of several consecutive predetermined points.
  • the process proceeds to S208.
  • a value that causes a failure is set on the basis of a database of experiments in which the device actually failed as in the case of the threshold A and the threshold time B.
  • the residual setting unit 11 calculates the residual in the current measured value from the slopes of a predetermined number of consecutive points up to the previous measured value. Next, the process proceeds to S209.
  • the residual is the current predicted value calculated from the slope of every predetermined number of points continuously measured up to the previous measurement value calculated by the vibration amplitude value analysis unit 10 by linear regression, and is actually measured. This is the difference from the current measured value.
  • the residual setting unit 11 sets a residual threshold.
  • a method for setting the residual threshold will be described. First, the specified multiples ⁇ ⁇ times are multiplied by the slope of each predetermined number of points that continue to the previous measurement value. The range of the predicted value calculated from the slope multiplied by the specified multiple is taken as the residual threshold.
  • the residual setting unit 11 advances the processing to S211 when the residual calculated with the data in the current measurement value exceeds the range of the residual threshold not less than the lower limit and not more than the upper limit. If the residual calculated with the current measurement value does not exceed the residual threshold, the process proceeds to S212.
  • the residual threshold is set to a specified multiple.
  • the absolute value of the residual threshold can be changed based on the slope of every predetermined number of points up to the previous measurement value.
  • the residual setting unit 11 does not proceed to S211 when two or more measured values that are out of the residual threshold range are continuously measured.
  • the residual threshold is set to a value that causes a failure based on a database of experiments in which the device actually failed.
  • the residual setting unit 11 excludes measurement values that are outside the residual threshold range. Then, the process proceeds to S212.
  • the vibration amplitude value analysis unit 10 ends the failure analysis of the amplitude value at the specific frequency f when a predetermined time has elapsed, and returns to the process of S201. Note that, when returning to the processing of S201, the contents stored in the memory can be reset. If the predetermined time has not elapsed, the vibration amplitude value analysis unit 10 advances the process to S213.
  • the signal determination unit 6 is observed with a value at which the slope of every predetermined number of consecutive points is equal to or greater than the determination threshold and exceeds the determination continuation count. If YES, the process proceeds to S214, and it is determined that a device failure can be predicted.
  • the signal determination unit 6 ends the failure analysis of the amplitude value at the specific frequency f when the above condition is not satisfied. Then, the process returns to S201. In addition, when returning to the process of S1, the content preserve
  • the operation of the failure prediction system 1 in the fourth embodiment can be switched in the order of operations in S212 and S213.
  • the residual setting unit 11 calculates the residual of the current measurement value from the slopes of several predetermined points up to the previous measurement value. Then, the residual setting unit 11 sets a residual threshold, and performs exclusion when the residual of the current measurement value exceeds the residual threshold.
  • the residual setting unit 11 sets a residual threshold, vibration caused by noise or the like can be excluded, and thus linear regression with a high correlation coefficient is possible.
  • the residual setting unit 11 does not exclude measured values when two or more measured values that are outside the range of the residual threshold are calculated continuously. Therefore, it is possible to avoid the exclusion of the measurement value due to vibrations that are likely to cause failure.
  • the fourth embodiment is configured by the entire system configuration diagram of FIG. 7 similar to the third embodiment.
  • the vibration measuring device 3 measures the frequency and vibration amplitude value of the device, and sends a signal of the measured frequency and vibration amplitude value to the signal conversion unit 4 of the signal processor 7.
  • S302 processing is advanced.
  • the signal conversion unit 4 and the vibration analysis start unit 9 perform a signal analysis start process similar to the operation in S202 of the third embodiment. Next, the process proceeds to S303.
  • the vibration amplitude value analysis unit 10 sets a lower limit value and an upper limit value of the amplitude value to be measured in order to exclude an abnormal point. Next, the process proceeds to S304.
  • the vibration amplitude value analysis unit 10 compares whether the measured value is within a range set between the lower limit value and the upper limit value. If the measured value exceeds the range, the process proceeds to S305. If the measured value does not exceed the range, the process proceeds to S306.
  • the vibration amplitude value analysis unit 10 excludes measurement exceeding the above range as abnormal data. Next, the process proceeds to S306.
  • the vibration amplitude value analysis unit 10 performs linear regression using the time as an independent variable for the amplitude value at a specific frequency f stored in the memory 8. Next, the process proceeds to S307.
  • the vibration amplitude value analysis unit 10 calculates the inclination for each predetermined number of consecutive points.
  • the process proceeds to S308.
  • a value that causes a failure is set on the basis of a database of experiments in which the device actually failed as in the case of the threshold A and the threshold time B.
  • the residual setting unit 11 calculates a predicted value to be measured next time from the slope of every predetermined number of consecutive points including the current measured value. Next, the process proceeds to S309.
  • the residual setting unit 11 calculates a residual threshold in the next measurement from the predicted value, as in the third embodiment. Next, the process proceeds to S310.
  • the residual setting unit 11 stores the predicted value and the residual threshold in the next measurement in the memory 8.
  • the residual setting unit 11 calculates a residual from the predicted value in the current measurement stored in the memory 8 at the previous measurement and the data in the current measured value. Next, the process proceeds to S312.
  • the residual setting unit 11 determines that the residual calculated with the current measurement value exceeds the range between the lower limit value and the upper limit value of the residual threshold value in the current measurement stored in the memory 8. , The process proceeds to S313. If the residual calculated with the current measurement value does not exceed the residual threshold range stored in the memory 8, the process proceeds to S314.
  • the residual threshold is set to a specified multiple.
  • the absolute value of the residual threshold can be changed based on the slope of every predetermined number of points up to the previous measurement value.
  • the residual setting unit 11 does not advance the process to S313 when two or more measured values that are outside the range of the residual threshold are continuously measured.
  • the residual setting unit 11 excludes measurement values that exceed the residual threshold. Then, the process proceeds to S314.
  • the vibration amplitude value analysis unit 10 ends the failure analysis of the amplitude value at the specific frequency f when a predetermined time elapses, and returns to the process of S301. Note that the content stored in the memory can be reset when returning to the processing of S301. If the predetermined time has not elapsed, the vibration amplitude value analysis unit 10 advances the process to S315.
  • step S315 the signal determination unit 6 is observed with a value at which the slope of every predetermined number of consecutive points is not less than the determination threshold and exceeds the determination continuation count. If YES in step S316, the process advances to step S316 to determine that a failure of the device can be predicted.
  • the operation of the failure prediction system 1 in the fourth embodiment can be switched in order of operations in S314 and S315.
  • 9A and 9B in the flowcharts of S308 to S310 indicated by A can use a CPU different from the CPU that operates the failure prediction system 1.
  • a CPU different from the CPU that operates the failure prediction system 1 can be used.
  • the residual setting unit 11 calculates a predicted value to be measured next time from the slopes of a predetermined number of consecutive points including the current measured value. Then, the residual setting unit 11 calculates a residual threshold in the next measurement from the predicted value, calculates the predicted value and the residual threshold, and stores them in the memory 8.
  • the failure prediction system 1 in the fourth embodiment calculates a residual that is a difference between the current measurement value and the prediction value stored in the memory 8. Then, by comparing the residual with the residual threshold stored in the memory 8, it can be determined whether the measured value exceeds the residual threshold.
  • noise or the like that is a measurement value outside the residual threshold is obtained. It is possible to quickly perform an operation process that excludes the caused vibration.
  • the failure prediction system 1 in the fourth embodiment performs the calculation of the prediction value and the residual threshold in the operations from S308 to S310 by using a CPU different from the CPU that operates the failure prediction system 1, thereby obtaining the prediction value.
  • the residual threshold value can be calculated quickly. As a result, it can be quickly determined whether the measured value is within the range determined by the residual threshold.
  • the failure prediction system 1 of the present invention is applied to the storage device 2 mounted on the computer device 13. Will be explained.
  • Fig. 10 shows the overall system configuration.
  • the failure prediction system 1 is located inside a computer device 13 that is a target of failure prediction, and includes a vibration measuring device 3, a signal processor 7, and a memory 8.
  • the computer device 13 includes a storage device 2 and a control circuit 12.
  • the signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6, each having the same connection relationship and function as in the fourth embodiment. Further, the vibration analysis starting unit 9, the vibration amplitude value analyzing unit 10, and the residual setting unit 11 included in the signal analyzing unit 5 have the same connection relationship as in the fourth embodiment.
  • the vibration measuring instrument 3 is attached to the storage device 2 of the computer device 13, and the vibration amplitude value analyzing unit 10 of the signal analyzing unit 5 is connected to the memory 8.
  • the signal determination unit 6 is connected to the control circuit 12 and transmits failure information to the control circuit 12 when it is determined that there is a failure.
  • the operation is the same as that of the first embodiment to the fourth embodiment.
  • the signal determination unit 6 predicts a failure, the signal determination unit 6 transmits failure information to the control circuit 12 of the computer device 13.
  • the failure prediction operation can be performed a plurality of times. After the signal determination unit 6 determines that the failure can be predicted, the control circuit 12 counts the number of times that the failure is predicted and the operation returns to the beginning.
  • control circuit 12 When the number of times that the control circuit 12 determines that a failure can be predicted is the first time, the control circuit 12 indicates warning information to the user of the computer device 13. The control circuit 12 moves the data to another device through the Internet or wiring if the number of times it is determined that the failure can be predicted is the second time, and if it is the third time, the computer device is assumed to be likely to fail. 13 is forcibly terminated.
  • the failure prediction system 1 in the fifth embodiment provides a failure prediction system 1 that does not deteriorate the performance of the normal operation of the computer device 13 and has high failure prediction accuracy, similarly to the effects up to the fourth embodiment. be able to.
  • the failure prediction system 1 can be stored in the computer device 13 to use the device memory 14 of the computer device 13 to store data based on failure prediction analysis. Therefore, the failure prediction system 1 does not need to provide a new memory, and can achieve the above-described effects with a low-cost and compact configuration.
  • the failure prediction system 1 can also provide an external memory 16 in the external device 15.
  • the user can set the external memory 16 having a necessary capacity according to the amount of data to be handled after using the computer device 13 in which the failure prediction system 1 has been introduced. Therefore, the user does not need to spend unnecessary costs, and can select the external memory 16 of the failure prediction system 1 according to the usage situation.
  • the number of times that the control circuit 12 determines that the failure can be predicted is counted, and according to the counted number, warning information is displayed to the user, data is transferred to another device, and the computer device 13 Forced termination is performed automatically. For this reason, when the user is not aware of the warning information, the number of times it is determined that a failure can be predicted is counted, and when the risk increases, processing such as data movement or forced termination of the computer device 13 is automatically performed. Data safety is ensured.
  • the failure prediction system 1 is located outside the computer device 13 that is a target of failure prediction, and is connected to the computer device 13 from the outside.
  • the failure prediction system 1 includes a vibration measuring instrument 3, a signal processor 7, and a memory 8.
  • the computer device 13 includes a storage device 2, a memory 8, and a control circuit 12.
  • the signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6, each having the same connection relationship and function as in the fifth embodiment. Further, the vibration analysis start unit 9, the vibration amplitude value analysis unit 10, and the residual setting unit 11 included in the signal analysis unit 5 have the same connection relationship as in the fifth embodiment.
  • failure prediction system 1 is the same as that of the fifth embodiment, and therefore will be omitted.
  • the failure prediction system 1 in the sixth embodiment can provide the failure prediction system 1 with high failure prediction accuracy without degrading the performance of the normal operation of the computer device 13, similarly to the effect of the fifth embodiment.
  • the failure prediction system 1 is connected to the computer device 13 from the outside. Therefore, a user who does not need the failure prediction system 1 can suppress an increase in price caused by mounting the failure prediction system 1 on the computer device 13 from the beginning. Then, after using the computer device 13, the user can introduce the failure prediction system 1 as required by the user.
  • the failure prediction system 1 can use the device memory 14 built in the computer device 13. Therefore, since the user does not need a new memory, the above-described effects can be achieved with a low-cost and compact configuration.
  • the failure prediction system 1 can also provide an external memory 16 in the external device 15.
  • the user can set the external memory 16 having a necessary capacity according to the amount of data handled by the user after introducing and using the failure prediction system 1 in the computer device 13. Therefore, the user does not need to spend unnecessary cost, and can select the external memory 16 of the failure prediction system 1 according to the usage situation.
  • the control circuit 12 counts the number of times it is determined that a failure can be predicted, and displays warning information to the user or data to other devices according to the counted number. And the forced termination of the computer device 13 are automatically performed. For this reason, when the user is not aware of the warning information, the number of times that it is determined that the failure can be predicted is counted, and when the risk increases, processing such as data movement or forced termination of the computer device 13 is automatically performed. As a result, the safety of the data is ensured.
  • Fig. 16 shows the overall system configuration.
  • the failure prediction system 1 is located inside a hard disk device 17 that is a target of failure prediction, and includes a vibration measuring device 3, a signal processor 7, and a memory 8.
  • the hard disk device 17 includes a hard disk control circuit 18.
  • the signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6, each having the same connection relationship and function as in the sixth embodiment. Further, the vibration analysis starting unit 9, the vibration amplitude value analyzing unit 10, and the residual setting unit 11 included in the signal analyzing unit 5 have the same connection relationship as in the sixth embodiment.
  • the vibration measuring instrument 3 is attached to the hard disk device 17, and the signal determination unit 6 is connected to the hard disk control circuit 18.
  • the signal determination unit 6 determines that a failure has occurred, the signal determination unit 6 transmits failure information to the hard disk control circuit 18.
  • failure prediction system 1 is the same as that in the sixth embodiment, and therefore will be omitted.
  • the vibration measuring instrument 3 measures the vibration of the hard disk device 17. If the signal determination unit 6 determines that a failure can be predicted, the signal determination unit 6 transmits failure information to the hard disk control circuit 18 of the hard disk device 17.
  • the failure prediction operation can be performed a plurality of times as in the sixth embodiment. After the signal determination unit 6 determines that the failure can be predicted, the hard disk control circuit 18 counts the number of times that the failure is predicted and the operation returns to the beginning.
  • the hard disk control circuit 18 transmits warning information to a device on which the hard disk device 17 is mounted according to the number of times it is determined that a failure can be predicted.
  • the failure prediction system 1 in the seventh embodiment is not limited to a specific computer device 13 by being incorporated in the hard disk device 17, and the failure prediction system 1 can be applied to various devices in which the hard disk device 17 is mounted. it can.
  • the failure prediction system 1 as shown in FIG. 17 can be stored in the hard disk device 17 to use the device memory 14 of the hard disk device 17 to store data by analyzing the failure prediction. Therefore, the failure prediction system 1 does not need to provide a new memory, and can achieve the above-described effects with a low-cost and compact configuration.
  • the hardware control circuit 12 counts the number of times it is determined that a failure can be predicted, and transmits warning information to a device on which the hard disk device 17 is mounted according to the count. Therefore, the device can recognize the degree of danger according to the number of transmitted information, and can perform processing for ensuring safety such as data movement.
  • the failure prediction system 1 is located inside a computer device 13 that is a target of failure prediction, and includes a vibration measuring device 3 and a signal processor 7.
  • the computer device 13 includes a storage device 2 and a control circuit 12.
  • the signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, a signal determination unit 6, and a memory 8.
  • the vibration measuring instrument 3 is attached to the storage device 2 of the computer device 13 and has a function of measuring the vibration of a device that is a target of failure prediction.
  • the signal converter 4 is connected to the vibration measuring instrument 3 and the signal analyzer 5 and has a function of converting an analog signal of vibration into a digital signal.
  • the signal analysis unit 5 includes a vibration analysis start unit 9, a vibration amplitude value analysis unit 10, and a residual setting unit 11.
  • the vibration analysis start unit 9 is connected to the signal conversion unit 4 and calculates the vibration spectrum of the vibration digital signal as seen from the frequency value of the amplitude value. When the vibration amplitude value exceeds a predetermined threshold value and continues for a predetermined time or more at a specific frequency f, analysis of the amplitude value is started.
  • the vibration amplitude value analysis unit 10 is connected to the vibration analysis start unit 9, and performs linear regression with the time as an independent variable for the vibration amplitude value, and calculates the slope of the line for each measurement time.
  • the residual setting unit 11 is connected to the vibration amplitude value analyzing unit 10 and has a function of calculating a residual threshold value from the slope for each measurement time and excluding the residual exceeding the residual threshold value.
  • the signal determination unit 6 is connected to the signal analysis unit 5 and the control circuit 12 of the computer device 13, and transmits failure information to the control circuit 12 when it is determined that there is a failure.
  • the vibration measuring instrument 3 uses the vibration sensor 19.
  • a piezoelectric acceleration sensor having a length of 7 mm, a width of 7 mm, and a height of 5 mm was installed at the center of the surface of the storage device 2.
  • the vibration measuring device 3 can obtain the same effect even if the vibration of the storage device 2 is measured using the acoustic microphone 20 instead of the vibration sensor 19.
  • the self-resonant frequency of the vibration sensor 19 used in this example is 50 kHz or more, and the frequency band to be measured is 1 Hz to 20 kHz.
  • the threshold value A of the amplitude value of the vibration spectrum to start measurement was set to 0.5 ⁇ m, and the threshold value B of the time for continuing the threshold value A was set to 10 seconds.
  • the amplitude data was measured at intervals of 5 seconds, and the slope by linear regression was calculated every 10 consecutive points, and the residual threshold was set to ⁇ 5 times.
  • the judgment conditions were such that the judgment threshold was set to 10 times the slope in linear regression with respect to the initial slope, and the number of judgment continuations was five.
  • the method, dimensions, installation location, measurement frequency band, self-resonant frequency, measurement start condition and determination condition of the vibration sensor 19 are not limited to this.
  • the vibration measuring instrument 3 measured the vibration at the center of the surface of the storage device 2 with the vibration sensor 19 and transmitted the amplified analog signal of the vibration to the signal conversion unit 4 of the signal processor 7.
  • the analog signal of vibration measured in the signal conversion unit 4 was converted into a digital signal and transmitted to the signal analysis unit 5.
  • the vibration analysis start unit 9 in the signal analysis unit 5 applied a discrete fast Fourier transform process to the vibration digital signal from the signal conversion unit 4 and analyzed the amplitude value of the vibration spectrum.
  • the frequency range of analysis was 1 Hz to 20 KHz.
  • the vibration analysis starting unit 9 saved the data in the memory 8.
  • the vibration analysis starting unit 9 observed that the amplitude value of vibration exceeded a predetermined threshold A of 0.5 ⁇ m at 200 Hz was observed for 10 seconds or more which is the threshold time B or longer and stored in the memory.
  • the vibration amplitude value analysis unit 10 started to analyze the amplitude value at 200 Hz, and stored the time change data of the amplitude value at 200 Hz in the memory 8.
  • the vibration amplitude value analysis unit 10 performed linear regression using time as an independent variable for the amplitude value at 200 Hz stored in the memory 8, and calculated the slope of the straight line every 10 consecutive points.
  • the residual setting unit 11 calculated the residual at each measurement time.
  • the residual setting unit 11 calculates a predicted value to be measured next from the slope obtained by linear regression at the last 10 consecutive points, and sets the range of the predicted value calculated by multiplying the slope by ⁇ 5 as a residual threshold.
  • the residual setting unit 11 stores the prediction value and residual threshold data in the memory 8.
  • the residual setting unit 11 calculates a residual from the measured value and the predicted value stored in the memory 8.
  • the residual setting unit 11 excludes measurement values whose residuals are outside the residual threshold range.
  • the threshold value for determination was set to 10 times the slope of linear regression at the first 10 points calculated after the start of amplitude value measurement.
  • the signal determination unit 6 determines whether the value of the slope at 10 consecutive measurement values exceeds the determination threshold.
  • the signal determination unit 6 was continuously observed five times or more in a state in which the slope of 10 consecutive points exceeded the determination threshold, 100 seconds after starting the measurement of the time variation of the amplitude value at 200 Hz.
  • the signal determination unit 6 determines that a failure of the storage device 2 is predicted, and transmits failure information to the control circuit 12 of the computer device 13.
  • FIG. 18 shows experimental results in this example and experimental results in which the failure prediction system of the present invention is applied to other failure factors.
  • the vibration amplitude of a specific frequency f caused by a failure is directly extracted and measured, and the transition of the amplitude value uses the residual to improve the accuracy of the regression line. Therefore, failure prediction of the storage device 2 can be realized with high accuracy.
  • vibration measurement unit 2 storage device 3: vibration measuring instrument 4: signal conversion unit 5: signal analysis unit 6: signal determination unit 7: signal processor 8: memory 9: vibration analysis start unit 10: vibration amplitude value analysis unit 11: Residual setting unit 12: Control circuit 13: Computer device 14: Device memory 15: External device 16: External memory 17: Hard disk device 18: Hard disk control circuit 19: Vibration sensor 20: Acoustic microphone

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)

Abstract

A　failure prediction system comprising a vibration measuring device which measures the vibration generated from a device, and a signal processor which predicts that there is a failure if the amplitude value of the vibration, measured by the vibration measuring device, has continued to exceed a predetermined threshold value for a predetermined amount of time or more. Further, the vibration measuring device measures the amplitude value of the vibration from the device at each frequency. Further, the signal processor predicts failures based on the change in the amplitude value over time. Here, the signal processor performs a linear regression on the amplitude value with time as the independent variable, calculates the slope of the line at each predetermined measurement time, then predicts failures in the device according to the trend in the slope of the line. Due to this, the system can suppress decrease in operation performance in a device without increasing how frequently failure prediction is performed.

Description

故障予測システムおよび電子機器、故障予測方法Failure prediction system, electronic device, and failure prediction method
　本発明は、故障予測システムに関し、特に電子機器等の故障予測システムに関する。 The present invention relates to a failure prediction system, and more particularly to a failure prediction system for electronic devices and the like.
　近年様々な電子機器が、私たちの生活における多くの場面で導入され利用されている。また電子化された情報は、電子機器に組み込まれることで膨大な量となっている。このような電子化された情報や多くの電子機器は、私たちの生活に欠かせないものとなっている。一方、電子機器は多数の機構部品が複雑に構成されているため、稼働中に機構部品が多数回動作によって劣化して破損してしまい故障する場合や、また突発的な衝撃により故障が発生してしまう場合がある。このような故障が起こると、利用者は電子機器が利用できなくなるだけでなく、大切な情報を消失してしまうことにもなりかねない。またこのような故障は、場合によっては私達の生活に経済的に大幅な損失をもたらしてしまうこともある。そのため故障予測システムの開発が、電子機器などの装置において進められている。 In recent years, various electronic devices have been introduced and used in many situations in our lives. Also, the information that has been digitized has become a huge amount by being incorporated into electronic equipment. Such electronic information and many electronic devices are indispensable for our lives. On the other hand, since many mechanical parts are complicatedly configured in electronic equipment, the mechanical parts may deteriorate and break down due to multiple operations during operation, or may fail due to sudden impacts. May end up. If such a failure occurs, the user may not only be able to use the electronic device, but also lose important information. Such failures can also cause significant economic losses to our lives in some cases. Therefore, the development of failure prediction systems is being promoted in devices such as electronic devices.
　このような故障予測システムに関連する文献として、特許文献１に、磁気ディスク装置における障害予測のシステムが記載されている。 As a document related to such a failure prediction system, Patent Document 1 describes a failure prediction system in a magnetic disk device.
　図１９を用いて、磁気ディスク装置の構成を説明する。図１９は、磁気ディスク装置の構成を示すブロック図である。同図に示すようにこの磁気ディスク装置１０１は、自動増幅調整（ＡＧＣ：Ａｕｔｏｍａｔｉｃ  Ｇａｉｎ  Ｃｏｎｔｒｏｌ）を備えるリードチャネル１0６と、メモリ１０９と、ディスク媒体１１５と、ヘッド監視部１１７と、障害予測部１１８とを少なくとも備える。 The configuration of the magnetic disk device will be described with reference to FIG. FIG. 19 is a block diagram showing the configuration of the magnetic disk device. As shown in the figure, the magnetic disk apparatus 101 includes a read channel 106 having automatic amplification adjustment (AGC), a memory 109, a disk medium 115, a head monitoring unit 117, and a failure prediction unit 118. At least.
　ＡＧＣとは、ディスク媒体１１５からサーボデータを読み出した際に、ヘッドから出力される信号を自動増幅する調節機構をいう。例えば、ヘッドが浮上してディスク媒体１１５から離れると、ヘッド先端に発生している磁界のディスク媒体１１５への作用が弱まることにより、弱まったヘッドからの出力信号を自動増幅される。障害予測部１１８は、AGCの信号出力強度が所定の範囲内、つまり正常に読み書きが行える状態時の増幅率であるか否かに基づいて、ヘッドの異常浮上を検出するものである。 AGC refers to an adjustment mechanism that automatically amplifies a signal output from the head when servo data is read from the disk medium 115. For example, when the head is lifted and separated from the disk medium 115, the magnetic field generated at the head tip is weakened on the disk medium 115, so that the output signal from the weakened head is automatically amplified. The failure prediction unit 118 detects the abnormal flying of the head based on whether or not the AGC signal output intensity is within a predetermined range, that is, whether or not the gain is in a state where reading and writing can be normally performed.
　次に特許文献１における障害予測の動作について説明する。ヘッド監視部１１７は、ヘッドの異常浮上回数を測定する一定のタイミングがくると、ディスク媒体１１５内に設けた一部の測定ゾーンについて、リードチャネル１１６内に備えられたＡＧＣの値を監視して測定する。 Next, the operation of failure prediction in Patent Document 1 will be described. The head monitoring unit 117 monitors the value of the AGC provided in the read channel 116 for a part of measurement zones provided in the disk medium 115 when a certain timing for measuring the number of abnormal flying of the head comes. taking measurement.
　障害予測部１１８は、ヘッド監視部１１７において測定をしたＡＧＣの値がディスク媒体１１５に記録されたＡＧＣ基準値よりも大きい場合には、ヘッドが異常浮上しているものとして検出する。そして障害予測部１１８は、ヘッドの異常浮上累積回数と、定期的に行われる測定時間区分（例えば、測定時の古い順から区分Ａ、Ｂ、ＣおよびＤ）ごとの異常浮上回数をメモリ１０９に保存する。障害予測部１１８は、メモリ９に保存されている測定データから、出荷時からの測定時間区分および異常浮上回数からなるグラフを作成する。そして障害予測部１１８は、異常浮上回数が増加傾向にあるか、または異常浮上回数の出荷時からの累積回数が閾値を超えていないかを算出する。 If the AGC value measured by the head monitoring unit 117 is larger than the AGC reference value recorded on the disk medium 115, the failure prediction unit 118 detects that the head is abnormally flying. Then, the failure prediction unit 118 stores in the memory 109 the cumulative number of abnormal levitations of the head and the number of abnormal levitations for each measurement time period (for example, classifications A, B, C, and D from the oldest at the time of measurement). save. The failure prediction unit 118 creates a graph including the measurement time segment from the time of shipment and the number of abnormal ascents from the measurement data stored in the memory 9. Then, the failure prediction unit 118 calculates whether the abnormal ascent number tends to increase or the accumulated number of abnormal ascent times from the shipment does not exceed the threshold value.
　障害予測部１１８は、異常浮上回数が増加傾向にある場合には、メモリ１０９に記録されている前回測定時のヘッドの異常浮上回数が増加傾向にあるか否か確認する。障害予測部１１８は、前回測定時もヘッドの異常浮上回数が増加傾向にある場合には、障害が発生する可能性が高いとして警告を発する。 The failure prediction unit 118 confirms whether or not the abnormal flying frequency of the head at the previous measurement recorded in the memory 109 is increasing when the abnormal flying frequency is increasing. The failure prediction unit 118 issues a warning that there is a high possibility that a failure will occur if the number of abnormal flying of the head tends to increase even during the previous measurement.
　障害予測部１１８は、異常浮上回数が増加傾向にない場合には、メモリ１０９に記録されている異常浮上回数の累積回数が閾値を超えているか否か確認する。障害予測部１１８は、異常浮上回数の累積回数が閾値を超えている場合には、障害が発生する可能性が高いとして警告を発する。 The failure prediction unit 118 checks whether or not the cumulative number of abnormal levitation times recorded in the memory 109 exceeds the threshold when the abnormal levitation frequency does not tend to increase. If the cumulative number of abnormal ascents exceeds the threshold, the failure predicting unit 118 issues a warning that there is a high possibility that a failure will occur.
　また他の故障予測システムに関連する文献として、特許文献２には軸受からのＡＥ（Ａｃｏｕｓｔｉｃ　Ｅｍｉｓｓｉｏｎ）を検出して異常を診断する、異常検出装置が記載さている。 Also, as a document related to other failure prediction systems, Patent Document 2 describes an abnormality detection device that diagnoses an abnormality by detecting AE (Acoustic Emission) from a bearing.
　ＡＥとは、材料が変形したりき裂が発生したりする際に、材料が内部に蓄えていたひずみエネルギーを弾性波として放出する現象のことである。 AE is a phenomenon in which strain energy stored in the material is released as an elastic wave when the material is deformed or a crack is generated.
　図２０を用いて特許文献２の軸受の異常検出装置の構成を説明する。異常検出装置は、ピークホールド回路２０６と、基準値発生器２０７と、比較器２０８と、ＡＤ変換器２１０と、ＣＰＵ(中央処理装置)２１１を備えている。 The configuration of the bearing abnormality detection device of Patent Document 2 will be described with reference to FIG. The abnormality detection apparatus includes a peak hold circuit 206, a reference value generator 207, a comparator 208, an AD converter 210, and a CPU (central processing unit) 211.
　次に、異常検出の動作について説明する。ピークホールド回路２０６は、ＡＥ信号を検出すると、このＡＥ信号のピーク値を検出して保持し、ＡＤ変換器２１０および比較器２０８に、ピーク値を出力する。 Next, the abnormality detection operation will be described. When the peak hold circuit 206 detects the AE signal, the peak hold circuit 206 detects and holds the peak value of the AE signal, and outputs the peak value to the AD converter 210 and the comparator 208.
　比較器２０８は、基準値発生器２０７が発生する基準値と、ピークホールド回路２０６からのピーク値とを比較し、ピーク値が基準値を越えたときに、トリガー信号をＡＤ変換器２１０とCPU２１１に出力する。 The comparator 208 compares the reference value generated by the reference value generator 207 with the peak value from the peak hold circuit 206. When the peak value exceeds the reference value, the trigger signal is sent to the AD converter 210 and the CPU 211. Output to.
　そしてＡＤ変換器２１０は、比較器２０８からトリガー信号を受けたときにＡＥ信号が処理可能な状態になり、ＡＥ信号をＡＤ変換する。 The AD converter 210 is ready to process the AE signal when it receives the trigger signal from the comparator 208, and AD converts the AE signal.
　CPU２１１も、比較器８からトリガー信号を受けたときにＡＥ信号が処理可能な状態になり、ＡＤ変換器２１０からデジタル化されたＡＥ信号を取り込んで、軸受の異常を診断する。したがってＡＤ変換器２１０とＣＰＵ２１１は、ＡＥ信号のピーク値が基準値を越えたときだけ、ＡＥ信号を信号処理する動作をする。 The CPU 211 is also in a state where the AE signal can be processed when receiving the trigger signal from the comparator 8, and takes the digitized AE signal from the AD converter 210 and diagnoses a bearing abnormality. Therefore, the AD converter 210 and the CPU 211 operate to process the AE signal only when the peak value of the AE signal exceeds the reference value.
　つまりＡＤ変換器２１０とＣＰＵ２１１は、ＡＥ信号のピーク値が基準値を下回っているときには、ＡＥ信号を信号処理する動作をしない。すなわち、ＡＤ変換器２１０とＣＰＵ２１１は、ＡＥ信号が軸受の異常診断に有効でない値のときには、ＡＥ信号を信号処理する動作を行わないため、通常動作性能を低下させず、また故障予測精度が高い故障予測システムを提供することができる。 That is, the AD converter 210 and the CPU 211 do not perform an operation to process the AE signal when the peak value of the AE signal is lower than the reference value. That is, since the AD converter 210 and the CPU 211 do not perform an operation for signal processing of the AE signal when the AE signal is a value that is not effective for bearing abnormality diagnosis, the normal operation performance is not deteriorated and the failure prediction accuracy is high. A failure prediction system can be provided.
特開2007-335013号公報Japanese Unexamined Patent Publication No. 2007-335013 特開平7-77459号公報JP-A-7-77459
　しかしながら上述のような特許文献１における障害予測システムでは、計測ごとに故障予測の増減の推移や比較を行うために、故障予測を行う頻度が高くなってしまう。そのため装置の演算処理機能に負荷がかかり、装置の通常動作に影響を与え性能を低下してしまう問題点があった。また特許文献２における異常検出装置では、周辺の一時的な機器雑音等に起因した振動、突発的に起こる環境振動などを誤ってエラーカウントしまい、正常にもかかわらず計測を開始してしまうおそれがあった。そのため、故障とは関係のない振動の場合においても、故障予測を行ってしまう。そのため、故障予測を行う頻度が高くなってしまい、演算処理に負荷がかかってしまう問題点があった。 However, in the failure prediction system in Patent Document 1 as described above, the frequency of failure prediction increases because a change or comparison of failure prediction is performed for each measurement. For this reason, there is a problem that a load is applied to the arithmetic processing function of the apparatus, affecting the normal operation of the apparatus and degrading the performance. In addition, in the abnormality detection device in Patent Document 2, vibrations caused by temporary peripheral equipment noises, sudden environmental vibrations, and the like are erroneously counted, and there is a possibility that measurement may be started despite normality. there were. Therefore, failure prediction is performed even in the case of vibrations unrelated to failure. For this reason, there is a problem that the frequency of failure prediction is increased and a load is imposed on the arithmetic processing.
　本発明の目的は、故障予測を行う頻度を高めることなく、装置の動作性能の低下を抑制する故障予測システムを提供することにある。 An object of the present invention is to provide a failure prediction system that suppresses a decrease in the operation performance of an apparatus without increasing the frequency of failure prediction.
　本発明の故障予測システムは、機器から発生する振動を計測する振動計測器と、前記振動計測器が計測した前記振動の振幅値が所定の閾値を越え、かつ所定の時間以上継続すると故障予測を行う信号処理器とを有している。 The failure prediction system according to the present invention includes a vibration measuring instrument that measures vibration generated from a device, and a failure prediction when an amplitude value of the vibration measured by the vibration measuring instrument exceeds a predetermined threshold and continues for a predetermined time or more. A signal processor to perform.
　本発明は、故障予測を行う頻度を高めることなく、装置の動作性能の低下を抑制することができる。 The present invention can suppress a decrease in the operating performance of the apparatus without increasing the frequency of failure prediction.
本発明の第１の実施形態に係る故障予測システムの構成図である。It is a lineblock diagram of a failure prediction system concerning a 1st embodiment of the present invention. 本発明の第１の実施形態に係る故障予測システムのフローチャートである。It is a flowchart of the failure prediction system which concerns on the 1st Embodiment of this invention. 振動の周波数スペクトル図である。It is a frequency spectrum figure of vibration. 本発明の第２の実施形態に係る故障予測システムの構成図である。It is a block diagram of the failure prediction system which concerns on the 2nd Embodiment of this invention. 本発明の第２の実施形態に係る故障予測システムのフローチャートである。It is a flowchart of the failure prediction system which concerns on the 2nd Embodiment of this invention. 振動の特定の周波数ｆにおける振幅推移図である。It is an amplitude transition diagram in the specific frequency f of a vibration. 本発明の第３の実施形態に係る故障予測システムの構成図である。It is a block diagram of the failure prediction system which concerns on the 3rd Embodiment of this invention. 本発明の第３の実施形態に係る故障予測システムのフローチャートである。It is a flowchart of the failure prediction system which concerns on the 3rd Embodiment of this invention. 本発明の第４の実施形態に係る故障予測システムのフローチャートである。It is a flowchart of the failure prediction system which concerns on the 4th Embodiment of this invention. 本発明の第４の実施形態に係る故障予測システムのフローチャートである。It is a flowchart of the failure prediction system which concerns on the 4th Embodiment of this invention. 本発明の第５の実施形態に係る故障予測システムをコンピュータ装置に内蔵した場合の構成図である。It is a block diagram at the time of incorporating the failure prediction system which concerns on the 5th Embodiment of this invention in the computer apparatus. 本発明の第５の実施形態に係る故障予測システムをコンピュータ装置に内蔵した場合、コンピュータ装置の装置メモリを用いた構成図である。When the failure prediction system according to the fifth embodiment of the present invention is built in a computer device, it is a configuration diagram using a device memory of the computer device. 本発明の第５の実施形態に係る故障予測システムをコンピュータ装置に内蔵した場合、外部装置の外部メモリを用いた構成図である。FIG. 10 is a configuration diagram using an external memory of an external device when a failure prediction system according to a fifth embodiment of the present invention is built in a computer device. 本発明の第６の実施形態に係る故障予測システムをコンピュータ装置に外付けした場合の構成図である。It is a block diagram at the time of attaching the failure prediction system which concerns on the 6th Embodiment of this invention externally to the computer apparatus. 本発明の第６の実施形態に係る故障予測システムをコンピュータ装置に外付けした場合、コンピュータ装置の装置メモリを用いた構成図である。When a failure prediction system according to a sixth embodiment of the present invention is externally attached to a computer device, it is a configuration diagram using a device memory of the computer device. 本発明の第６の実施形態に係る故障予測システムをコンピュータ装置に外付けした場合、外部装置の外部メモリを用いた構成図である。When the failure prediction system according to the sixth embodiment of the present invention is externally attached to a computer device, it is a configuration diagram using an external memory of an external device. 本発明の第７の実施形態に係る故障予測システムをハードディスク装置に内蔵した場合の構成図である。It is a block diagram at the time of incorporating the failure prediction system which concerns on the 7th Embodiment of this invention in a hard disk drive. 本発明の第７の実施形態に係る故障予測システムをハードディスク装置に内蔵した場合、ハードディスク装置の装置メモリを用いた構成図である。When the failure prediction system according to the seventh embodiment of the present invention is built in a hard disk device, it is a configuration diagram using a device memory of the hard disk device. 実施例に係る記憶装置の故障予測結果である。It is a failure prediction result of the memory | storage device based on an Example. 特許文献１における磁気ディスク装置の構成図である。1 is a configuration diagram of a magnetic disk device in Patent Document 1. FIG. 特許文献２における軸受の異常検出装置の構成図である。10 is a configuration diagram of a bearing abnormality detection device in Patent Document 2. FIG.
　[第１の実施形態]本発明の第１の実施形態を、図面を用いて詳細に説明する。 [First Embodiment] A first embodiment of the present invention will be described in detail with reference to the drawings.
　[構成の説明]図１に本発明の故障予測システム１の構成図を示す。本発明の故障予測システム１は、振動計測器３と信号処理器７とメモリ８を備えている。 [Description of Configuration] FIG. 1 shows a configuration diagram of the failure prediction system 1 of the present invention. The failure prediction system 1 of the present invention includes a vibration measuring instrument 3, a signal processor 7, and a memory 8.
　振動計測器３は、故障予測の対象となる機器の振動を計測する機能を有し、図示していない機器に取り付けられている。機器とはハードディスク装置のように振動を発生するものを備えた電子機器のことをいう。 The vibration measuring instrument 3 has a function of measuring the vibration of a device that is a target of failure prediction, and is attached to a device that is not shown. A device refers to an electronic device equipped with a device that generates vibration, such as a hard disk device.
　信号処理器７は、信号変換部４と信号解析部５と信号判定部６を備えている。 The signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6.
　信号変換部４は、振動計測器３と信号解析部５に接続され、振動のアナログ信号をデジタル信号へ変換する機能を有する。 The signal conversion unit 4 is connected to the vibration measuring instrument 3 and the signal analysis unit 5 and has a function of converting an analog signal of vibration into a digital signal.
　信号解析部５は、信号変換部４と接続しており、振動のデジタル信号に対して、振幅値の大きさを周波数の値から見た振動スペクトルにより計測を行う。振動スペクトルとは、図３のように縦軸を振動の振幅値、横軸を周波数としたグラフのことである。そして信号解析部５は、特定の周波数ｆにおいて、振動の振幅値が所定の閾値を越え、かつ所定の時間以上継続すると、特定の周波数ｆにおける振幅値の時間変化について解析をする。 The signal analysis unit 5 is connected to the signal conversion unit 4 and measures the amplitude value of the vibration digital signal based on the vibration spectrum viewed from the frequency value. As shown in FIG. 3, the vibration spectrum is a graph with the vertical axis representing the vibration amplitude value and the horizontal axis representing the frequency. When the vibration amplitude value exceeds a predetermined threshold at a specific frequency f and continues for a predetermined time or longer, the signal analysis unit 5 analyzes the time change of the amplitude value at the specific frequency f.
　信号判定部６は、信号解析部５と接続しており、特定の周波数ｆにおける振動の振幅値の時間変化に基づいて、故障予測を行う機能を有する。 The signal determination unit 6 is connected to the signal analysis unit 5 and has a function of performing failure prediction based on a temporal change in the amplitude value of vibration at a specific frequency f.
　メモリ８は、信号解析部５と接続している。 The memory 8 is connected to the signal analysis unit 5.
　[動作の説明]次に、本発明の故障予測システム１の動作を図２のフローチャートを用いて説明する。 [Description of Operation] Next, the operation of the failure prediction system 1 of the present invention will be described with reference to the flowchart of FIG.
　Ｓ１において、振動計測器３は、機器の周波数と振動振幅値とを計測し、計測した周波数と振動振幅値の信号を信号処理器７の信号変換部４へ送る。次にＳ２処理を進める。 In S <b> 1, the vibration measuring device 3 measures the frequency and vibration amplitude value of the device, and sends a signal of the measured frequency and vibration amplitude value to the signal conversion unit 4 of the signal processor 7. Next, S2 processing is advanced.
　Ｓ２において、信号変換部４は、振動計測器３からの振動のアナログ信号をデジタル信号へ変換し、信号解析部５に送信する。次にＳ３に処理を進める。 In S <b> 2, the signal conversion unit 4 converts an analog signal of vibration from the vibration measuring instrument 3 into a digital signal and transmits the digital signal to the signal analysis unit 5. Next, the process proceeds to S3.
　Ｓ３において、信号解析部５は、信号変換部４からの振動のデジタル信号に対して、振幅値の大きさを周波数の値から見た振動スペクトルにより解析を行う。次にＳ４に処理を進める。 In S <b> 3, the signal analysis unit 5 analyzes the vibration digital signal from the signal conversion unit 4 using a vibration spectrum in which the magnitude of the amplitude value is viewed from the frequency value. Next, the process proceeds to S4.
　Ｓ４において、信号解析部５は、図３に示すように例えば特定の周波数ｆにおける振動の振幅値が所定の閾値Ａを超えた場合、次にＳ５に処理を進める。振動の振幅値が所定の閾値Ａを超えない場合は、Ｓ１の処理に戻る。 In S4, when the amplitude value of vibration at a specific frequency f exceeds a predetermined threshold A as shown in FIG. 3, for example, the signal analysis unit 5 proceeds to S5. When the amplitude value of vibration does not exceed the predetermined threshold A, the process returns to S1.
　Ｓ５において、信号解析部５は振動の振幅値が所定の閾値Ａを超えたデータをメモリ８に保存する。 In S5, the signal analysis unit 5 stores data in which the amplitude value of vibration exceeds a predetermined threshold A in the memory 8.
　Ｓ６において、信号解析部５は特定の周波数ｆにおいて振動の振幅値が所定の閾値Ａを超えた状態で、所定の継続時間である閾値時間Ｂ以上経過すると、次にＳ７に処理を進める。振動の振幅値が閾値Ａを越えた状態で、閾値時間Ｂ以上継続して経過しなかった場合、Ｓ１の処理に戻る。なお、Ｓ１の処理に戻る際に、メモリに保存された内容をリセットすることもできる。 In S6, when the amplitude value of the vibration exceeds a predetermined threshold A at a specific frequency f and the threshold time B, which is a predetermined duration, elapses, the signal analysis unit 5 advances the process to S7. If the vibration amplitude value exceeds the threshold A and does not continue for more than the threshold time B, the process returns to S1. In addition, when returning to the process of S1, the content preserve | saved at memory can also be reset.
　ここで振動スペクトルの解析における振幅値の閾値Ａは、故障の要因となる振動に対して漏れを少なくなるように、かつ故障とは関係のない振動の故障予測の解析を始めてしまい、機器の通常の動作に影響を与えてしまうことのないように設定される。 Here, the threshold value A of the amplitude value in the analysis of the vibration spectrum starts the analysis of the failure prediction of the vibration not related to the failure so as to reduce the leakage with respect to the vibration that causes the failure, and the normal of the device. It is set so as not to affect the operation of.
　同様に振幅値の閾値Ａを超えた状態で継続する閾値時間Ｂについても、故障の要因となる振動に対して漏れを少なくなるように、かつ閾値Ｂについても、故障とは関係のない振動の故障予測の解析を始めてしまい、機器の通常の動作に影響を与えてしまうことのないように設定される。 Similarly, with respect to the threshold time B that continues in a state in which the amplitude value exceeds the threshold value A, the leakage of the vibration that causes the failure is reduced, and the threshold value B also has vibrations that are not related to the failure. It is set so that failure prediction analysis is started and normal operation of the device is not affected.
　そこで振幅値の閾値Ａと、閾値Ａを超えた状態の継続時間の閾値時間Ｂは、機器の故障要因に応じて発生する振動に基づいて決定する。実際に機器が故障に至った際の振動スペクトルの振幅値や、振幅値が一定の閾値をこえた状態で継続する時間を実験から周波数ごとに測定し、データベース化した。このデータベースに基づいて、故障の要因となる閾値Ａと閾値時間Ｂは、設定される。なお周波数に応じて故障の要因は異なるため、閾値Ａと閾値時間Ｂは周波数に応じて設定される。 Therefore, the threshold value A of the amplitude value and the threshold time B of the duration time in which the threshold value A is exceeded are determined based on vibrations generated according to the failure factor of the device. The amplitude value of the vibration spectrum when the device actually broke down and the duration of the amplitude value exceeding a certain threshold were measured from the experiment for each frequency and compiled into a database. Based on this database, a threshold A and a threshold time B that cause failure are set. In addition, since the cause of failure differs according to the frequency, the threshold A and the threshold time B are set according to the frequency.
　Ｓ７において、信号判定部６は特定の周波数ｆにおける振動振幅値の時間変化に基づいて、故障予測を開始する。 In S7, the signal determination unit 6 starts failure prediction based on the temporal change of the vibration amplitude value at the specific frequency f.
　[効果の説明]第１の実施形態における効果について説明する。第１の実施形態における故障予測システム１は、通常動作では図３のように振動解析開始部９は振動スペクトルの振幅値が閾値Ａを超えた値で、所定の継続時間である閾値時間Ｂ以上継続しているかのみモニタリングしている。つまり故障予測システム１は、故障予測を常に行っているのではなく、第一段階として故障への進行に伴う特徴的な振動を直接抽出する監視を行っている。 [Explanation of Effects] Effects in the first embodiment will be described. In normal operation, the failure prediction system 1 in the first embodiment is such that the vibration analysis starting unit 9 is a value that the amplitude value of the vibration spectrum exceeds the threshold value A as shown in FIG. It monitors only whether it continues. That is, the failure prediction system 1 does not always perform failure prediction, but performs monitoring to directly extract characteristic vibrations accompanying the progress to failure as a first step.
　これにより故障予測システム１は、故障に至る可能性の高い振動の特定周波数ｆを抽出することで、故障予測を行う頻度を抑制することが出来る。そのため故障予測システム１は、故障の進行に関与しない機器の振動のデータの取得や解析を行う必要がなく、突発的な振動と故障の進行に伴う振動を区別できる。その結果、故障予測システム１は、正常にも関わらず故障予測の解析を行うことで、対象となる機器の性能が低下することを抑制でき、機器の通常動作における演算処理などの性能を低下させず、負荷を抑えることができる。 Thus, the failure prediction system 1 can suppress the frequency of failure prediction by extracting the specific frequency f of vibration that is likely to cause a failure. Therefore, the failure prediction system 1 does not need to acquire and analyze vibration data of devices that are not involved in the progress of the failure, and can distinguish between sudden vibration and vibration accompanying the progress of the failure. As a result, the failure prediction system 1 can suppress a decrease in the performance of the target device by analyzing the failure prediction in spite of normality, thereby reducing the performance of arithmetic processing in the normal operation of the device. Therefore, the load can be suppressed.
　[第２の実施形態]本発明の第２の実施形態を、図面を用いて詳細に説明する。 [Second Embodiment] A second embodiment of the present invention will be described in detail with reference to the drawings.
　[構成の説明]図４にシステム全体構成図を示す。故障予測システム１は、第１の実施形態と同様に故信号計測器３と信号処理器７とメモリ８を備えている。また信号処理器７は、信号変換部４と信号解析部５と信号判定部６を備えている。各ブロックの接続関係についても、第１の実施形態と同様の接続関係を有している。 [Description of configuration] Fig. 4 shows the overall system configuration. The failure prediction system 1 includes a late signal measuring device 3, a signal processor 7, and a memory 8 as in the first embodiment. The signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6. The connection relationship between the blocks is the same as that in the first embodiment.
　第２の実施形態と第１の実施形態との相違点として、信号解析部５は振動解析開始部９と振動振幅値解析部１０を有している。 As a difference between the second embodiment and the first embodiment, the signal analysis unit 5 includes a vibration analysis start unit 9 and a vibration amplitude value analysis unit 10.
　振動解析開始部９は、信号変換部４と接続しており、振動のデジタル信号に対して、振幅値の大きさを周波数の値から見た振動スペクトルにより計測を行う。そして特定の周波数ｆにおいて、振動の振幅値が所定の閾値を越え、かつ所定の時間以上継続すると、特定の周波数ｆにおける振動の振幅値の時間変化について解析を開始する。 The vibration analysis start unit 9 is connected to the signal conversion unit 4 and measures the amplitude value of the vibration digital signal from the vibration spectrum as seen from the frequency value. When the vibration amplitude value exceeds a predetermined threshold at a specific frequency f and continues for a predetermined time or longer, an analysis is started for a temporal change in the vibration amplitude value at the specific frequency f.
　振動振幅値解析部１０は、振動解析開始部９と接続しており、振動の振幅値について時間を独立変数とした直線回帰を行い、所定の計測時間ごとに直線の傾きを算出する。 The vibration amplitude value analysis unit 10 is connected to the vibration analysis start unit 9, and performs linear regression with the time as an independent variable for the vibration amplitude value, and calculates the slope of the line every predetermined measurement time.
　[動作の説明]次に、第２の実施形態における故障予測システム１の動作を図５のフローチャートを用いて説明する。 [Description of Operation] Next, the operation of the failure prediction system 1 in the second embodiment will be described with reference to the flowchart of FIG.
　Ｓ１０１において、振動計測器３は、機器の周波数と振動振幅値とを計測し、計測した周波数と振動振幅値の信号を信号処理器７の信号変換部４へ送る。次にＳ１０２処理を進める。 In S <b> 101, the vibration measuring device 3 measures the frequency and vibration amplitude value of the device, and sends a signal of the measured frequency and vibration amplitude value to the signal conversion unit 4 of the signal processor 7. Next, S102 processing is advanced.
　Ｓ１０２における動作は、第１の実施形態のＳ２からＳ６における動作と同様であるため、故障予測解析処理として１つのステップとして示した。なお、第２の実施例では、Ｓ１０２におけるＳ３からＳ６に該当する動作は、信号解析部５における信号解析開始部９が処理を行う。そして信号解析開始部９が、特定の周波数ｆにおいて振動の振幅値が所定の閾値Ａを超えた状態で、所定の継続時間である閾値時間Ｂ以上経過すると、次にＳ１０３に処理を進める。 Since the operation in S102 is the same as the operation in S2 to S6 of the first embodiment, it is shown as one step as the failure prediction analysis process. In the second embodiment, the signal analysis start unit 9 in the signal analysis unit 5 performs the processes corresponding to S3 to S6 in S102. Then, when the signal analysis start unit 9 has exceeded the predetermined threshold A at the specific frequency f and the threshold time B, which is a predetermined duration, has elapsed, the process proceeds to S103.
　Ｓ１０３において、振動振幅値解析部１０は異常点を除外するために測定する振幅値の下限値と上限値を設定する。次にＳ１０４に処理を進める。 In S103, the vibration amplitude value analysis unit 10 sets a lower limit value and an upper limit value of an amplitude value to be measured in order to exclude an abnormal point. Next, the process proceeds to S104.
　Ｓ１０４において、振動振幅値解析部１０は、測定した振幅値が下限値以上、上限値以下で設定される範囲以内であるか比較を行う。振幅値が範囲を超えた場合はＳ１０５に処理を進める。測定値が範囲を超えない場合は、Ｓ１０６に処理を進める。 In S104, the vibration amplitude value analysis unit 10 compares whether the measured amplitude value is within a range set between the lower limit value and the upper limit value. If the amplitude value exceeds the range, the process proceeds to S105. If the measured value does not exceed the range, the process proceeds to S106.
　Ｓ１０５において、振動振幅値解析部１０は、上記の範囲を超えた振幅値を異常データとして除外する。次にＳ１０６に処理を進める。 In S105, the vibration amplitude value analysis unit 10 excludes amplitude values exceeding the above range as abnormal data. Next, the process proceeds to S106.
　Ｓ１０６において、振動振幅値解析部１０は、メモリ８に保存されている特定の周波数ｆにおける振幅値について時間を独立変数とした直線回帰を行う。次にＳ１０７に処理を進める。 In S106, the vibration amplitude value analysis unit 10 performs linear regression using the time as an independent variable for the amplitude value at a specific frequency f stored in the memory 8. Next, the process proceeds to S107.
　Ｓ１０７において、振動振幅値解析部１０は、連続する所定の数点ごとの傾きを算出する。次にＳ１０８に処理を進める。所定の数点とは、閾値Ａや閾値時間Ｂと同様に実際に機器が故障に至った実験のデータベースに基づいて、故障の要因となる値を設定する。 In S107, the vibration amplitude value analysis unit 10 calculates the inclination for each predetermined number of consecutive points. Next, the process proceeds to S108. As the predetermined number of points, a value that causes a failure is set on the basis of a database of experiments in which the device actually failed as in the case of the threshold A and the threshold time B.
　Ｓ１０８において、振動振幅値解析部１０は、所定の時間経過すると特定の周波数ｆにおける振幅値の故障解析を終了し、Ｓ１０１の処理に戻る。なお、Ｓ１０１の処理に戻る際に、メモリに保存された内容をリセットすることもできる。振動振幅値解析部１０は、所定の時間経過していなければ、Ｓ１０９に処理を進める。 In S108, the vibration amplitude value analysis unit 10 ends the failure analysis of the amplitude value at the specific frequency f when a predetermined time has elapsed, and returns to the process of S101. Note that when returning to the processing of S101, the content stored in the memory can be reset. If the predetermined time has not elapsed, the vibration amplitude value analysis unit 10 advances the process to S109.
　Ｓ１０９において、信号判定部６は、判定用閾値と判定用継続回数を各々設定し、直線回帰の傾きの推移から故障判定を行う。図６のように直線回帰による傾きの解析は連続する所定の数点ごとに算出する。 In S109, the signal determination unit 6 sets a determination threshold value and a determination continuation number, and performs a failure determination from the transition of the slope of the linear regression. As shown in FIG. 6, the slope analysis by linear regression is calculated for every predetermined number of consecutive points.
　そしてＳ１０９において、信号判定部６は、振動振幅値解析部１０が算出した連続する所定の数点ごとの傾きが、判定用閾値以上でかつ判定用継続回数を越えた値で観測された場合、Ｓ１１０に処理を進め機器の故障が予測できると判断する。 In S109, the signal determination unit 6 observes the slope of each predetermined number of points calculated by the vibration amplitude value analysis unit 10 at a value that is equal to or greater than the determination threshold and exceeds the determination continuation count. The process proceeds to S110 and it is determined that a failure of the device can be predicted.
　しかし、Ｓ１０９において、信号判定部６は上記の条件を満たさない場合は、特定の周波数ｆにおける振幅値の故障解析を終了する。そして、Ｓ１０１の処理に戻る。なお、Ｓ１０１の処理に戻る際に、メモリに保存された内容をリセットすることもできる。 However, in S109, when the signal determination unit 6 does not satisfy the above condition, the failure analysis of the amplitude value at the specific frequency f ends. Then, the process returns to S101. Note that when returning to the processing of S101, the content stored in the memory can be reset.
　判定用閾値は、特定の周波数ｆにおける振幅値を計測開始後、最初に算出した所定の数点における傾きａ１の指定数倍α倍に設定される。 The determination threshold value is set to a specified number of times α times the slope a1 calculated at a predetermined number of points first calculated after the amplitude value at a specific frequency f is started.
　また判定用継続回数は、測定したデータの連続する所定の数点ごとの傾きが判定用閾値を超えた値で、連続して発生する回数とする。 Also, the judgment continuation count is a value at which the slope of every predetermined number of consecutive measured data exceeds the judgment threshold value and is continuously generated.
　なお、閾値Ａや閾値時間Ｂと同様に、判定用閾値、判定用継続回数は、実際に機器が故障に至った実験のデータベースに基づいて、故障の要因となる値を設定する。 Note that, similarly to the threshold A and the threshold time B, the determination threshold and the determination continuation number are set to values that cause failure based on an experimental database in which the device actually failed.
　[効果の説明]第２の実施形態における効果について説明する。第２の実施形態における故障予測システム１は、振動解析開始部９が、特定の周波数ｆにおいて振動の振幅値が閾値Ａを超えた状態で、閾値時間Ｂ以上継続すると、機器が故障となる可能性が高いとして故障予測の解析を始める。そして振動振幅値解析部１０は、特定の周波数ｆにおける振動の振幅値について時間を独立変数とした直線回帰を行う。 [Explanation of Effects] Effects in the second embodiment will be described. In the failure prediction system 1 according to the second embodiment, when the vibration analysis start unit 9 continues for a threshold time B or more in a state where the amplitude value of vibration exceeds the threshold A at a specific frequency f, a device may fail. The failure prediction analysis is started because of the high probability. Then, the vibration amplitude value analysis unit 10 performs linear regression using the time as an independent variable for the vibration amplitude value at the specific frequency f.
　第２の実施形態における故障予測システム１は、信号判定部６が故障の進行に関与する振動に対して、判定用閾値と判定用継続回数を各々設定し、直線回帰の傾きの推移から故障判定を行うことで、精度の高い故障予測を行うことができる。 In the failure prediction system 1 according to the second embodiment, the signal determination unit 6 sets a determination threshold value and a determination continuation number for vibrations involved in the progress of the failure, and determines failure from the transition of the slope of linear regression. By performing this, it is possible to perform failure prediction with high accuracy.
　また、振動振幅値解析部１０は異常点を除外するために測定する振幅値の上限と下限の閾値を設定すること、簡素な算出処理によって雑音等に起因した振動を除外できる。 Also, the vibration amplitude value analysis unit 10 can exclude vibrations caused by noise or the like by setting the upper and lower thresholds of the amplitude value to be measured in order to exclude abnormal points, and by a simple calculation process.
　[第３の実施形態]本発明の第３の実施形態を、図面を用いて詳細に説明する。
　[構成の説明]図７にシステム全体構成図を示す。故障予測システム１は、第２の実施形態と同様に故信号計測器３と信号処理器７とメモリ８を備えている。また信号処理器７は、信号変換部４と信号解析部５と信号判定部６を備え、信号解析部５は振動解析開始部９と振動振幅値解析部１０を有している。各ブロックの接続関係についても、第２の実施形態と同様の接続関係を有している。 [Third Embodiment] A third embodiment of the present invention will be described in detail with reference to the drawings.
[Explanation of Configuration] FIG. 7 shows an overall system configuration diagram. The failure prediction system 1 includes a late signal measuring device 3, a signal processor 7, and a memory 8 as in the second embodiment. The signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6, and the signal analysis unit 5 includes a vibration analysis start unit 9 and a vibration amplitude value analysis unit 10. The connection relationship between the blocks is the same as that in the second embodiment.
　第３の実施形態と第２の実施形態との相違点として、信号解析部５は残差設定部１１を有している。残差設定部１１は、振動振幅値解析部１０と信号判定部６と接続しており、所定の計測時間ごとの傾きから残差閾値を計算し、残差閾値を越えた残差を除外する機能を有する。 As a difference between the third embodiment and the second embodiment, the signal analysis unit 5 has a residual setting unit 11. The residual setting unit 11 is connected to the vibration amplitude value analysis unit 10 and the signal determination unit 6, calculates a residual threshold value from a slope for each predetermined measurement time, and excludes residuals exceeding the residual threshold value. It has a function.
　[動作の説明]次に、第３の実施形態における故障予測システム１の動作を図８のフローチャートを用いて説明する。 [Description of Operation] Next, the operation of the failure prediction system 1 in the third embodiment will be described with reference to the flowchart of FIG.
　Ｓ２０１において、振動計測器３は、機器の周波数と振動振幅値とを計測し、計測した周波数と振動振幅値の信号を信号処理器７の信号変換部４へ送る。次にＳ２０２処理を進める。 In S201, the vibration measuring device 3 measures the frequency and vibration amplitude value of the device, and sends a signal of the measured frequency and vibration amplitude value to the signal conversion unit 4 of the signal processor 7. Next, the process of S202 is advanced.
　Ｓ２０２において、信号変換部４と振動解析開始部９は第２の実施形態のＳ１０２における動作と同様の信号解析開始処理を行う。次に、Ｓ２０３に処理を進める。 In S202, the signal conversion unit 4 and the vibration analysis start unit 9 perform a signal analysis start process similar to the operation in S102 of the second embodiment. Next, the process proceeds to S203.
　Ｓ２０３において、振動振幅値解析部１０は異常点を除外するために測定する振幅値の下限値と上限値を設定する。次にＳ２０４に処理を進める。 In S203, the vibration amplitude value analysis unit 10 sets a lower limit value and an upper limit value of the amplitude value to be measured in order to exclude abnormal points. Next, the process proceeds to S204.
　Ｓ２０４において、振動振幅値解析部１０は、測定した振幅値が下限値以上、上限値以下で設定される範囲以内であるか比較を行う。振幅値が範囲を超えた場合はＳ２０５に処理を進める。測定値が範囲を超えない場合は、Ｓ２０６に処理を進める。 In S204, the vibration amplitude value analysis unit 10 compares whether the measured amplitude value is within a range set between the lower limit value and the upper limit value. If the amplitude value exceeds the range, the process proceeds to S205. If the measured value does not exceed the range, the process proceeds to S206.
　Ｓ２０５において、振動振幅値解析部１０は、上記の範囲を超えた測定値を異常データとして除外する。次にＳ２０６に処理を進める。 In S205, the vibration amplitude value analysis unit 10 excludes measurement values exceeding the above range as abnormal data. Next, the process proceeds to S206.
　Ｓ２０６において、振動振幅値解析部１０は、メモリ８に保存されている特定の周波数ｆにおける振幅値について時間を独立変数とした直線回帰を行う。次にＳ２０７に処理を進める。 In S206, the vibration amplitude value analysis unit 10 performs linear regression using the time as an independent variable for the amplitude value at a specific frequency f stored in the memory 8. Next, the process proceeds to S207.
　Ｓ２０７において、振動振幅値解析部１０は、連続する所定の数点ごとの傾きを算出する。次にＳ２０８に処理を進める。所定の数点とは、閾値Ａや閾値時間Ｂと同様に実際に機器が故障に至った実験のデータベースに基づいて、故障の要因となる値を設定する。 In S207, the vibration amplitude value analysis unit 10 calculates the inclination for each of several consecutive predetermined points. Next, the process proceeds to S208. As the predetermined number of points, a value that causes a failure is set on the basis of a database of experiments in which the device actually failed as in the case of the threshold A and the threshold time B.
　Ｓ２０８において、残差設定部１１は前回の測定値までの連続する所定の数点ごとの傾きから、今回の測定値における残差を算出する。次にＳ２０９に処理を進める。 In S208, the residual setting unit 11 calculates the residual in the current measured value from the slopes of a predetermined number of consecutive points up to the previous measured value. Next, the process proceeds to S209.
　ここで残差について説明する。残差とは、図６に示すとおり振動振幅値解析部１０が直線回帰により算出した前回の測定値まで連続する所定の数点ごとの傾きから算出された今回の予測値と、実際に測定された今回の測定値との差である。 Here we explain the residual. As shown in FIG. 6, the residual is the current predicted value calculated from the slope of every predetermined number of points continuously measured up to the previous measurement value calculated by the vibration amplitude value analysis unit 10 by linear regression, and is actually measured. This is the difference from the current measured value.
　Ｓ２０９において、残差設定部１１は残差閾値を設定する。残差閾値の設定方法について説明する。まず、前回の測定値まで連続する所定の数点ごとの傾きに指定倍数±γ倍を乗じる。指定倍数を乗じた傾きから算出された予測値の範囲を残差閾値とする。 In S209, the residual setting unit 11 sets a residual threshold. A method for setting the residual threshold will be described. First, the specified multiples ± γ times are multiplied by the slope of each predetermined number of points that continue to the previous measurement value. The range of the predicted value calculated from the slope multiplied by the specified multiple is taken as the residual threshold.
　Ｓ２１０において、残差設定部１１は、今回の測定値におけるデータで算出された残差が残差閾値の下限値以上、上限値以下の範囲を越えた場合は、Ｓ２１１に処理を進める。今回の測定値で算出された残差が残差閾値を越えない場合は、Ｓ２１２に処理を進める。 In S210, the residual setting unit 11 advances the processing to S211 when the residual calculated with the data in the current measurement value exceeds the range of the residual threshold not less than the lower limit and not more than the upper limit. If the residual calculated with the current measurement value does not exceed the residual threshold, the process proceeds to S212.
　なお、上記において残差閾値は指定倍数とされているが、前回の測定値までの連続する所定の数点ごとの傾きに基づいて残差閾値を絶対値を変動することもできる。ただし、Ｓ２１０において残差設定部１１は、残差閾値の範囲外である測定値が２点以上連続して計測された場合は、Ｓ２１１に処理を進めない。 In the above description, the residual threshold is set to a specified multiple. However, the absolute value of the residual threshold can be changed based on the slope of every predetermined number of points up to the previous measurement value. However, in S210, the residual setting unit 11 does not proceed to S211 when two or more measured values that are out of the residual threshold range are continuously measured.
　閾値Ａや閾値時間Ｂと同様に、残差閾値は、実際に機器が故障に至った実験のデータベースに基づいて、故障の要因となる値を設定する。 Like the threshold A and the threshold time B, the residual threshold is set to a value that causes a failure based on a database of experiments in which the device actually failed.
　Ｓ２１１において、残差設定部１１は、残差閾値の範囲外である測定値を除外する。そしてＳ２１２に処理を進める。 In S211, the residual setting unit 11 excludes measurement values that are outside the residual threshold range. Then, the process proceeds to S212.
　Ｓ２１２において、振動振幅値解析部１０は、所定の時間経過すると特定の周波数ｆにおける振幅値の故障解析を終了し、Ｓ２０１の処理に戻る。なお、Ｓ２０１の処理に戻る際に、メモリに保存された内容をリセットすることもできる。振動振幅値解析部１０は、所定の時間経過していなければ、Ｓ２１３に処理を進める。 In S212, the vibration amplitude value analysis unit 10 ends the failure analysis of the amplitude value at the specific frequency f when a predetermined time has elapsed, and returns to the process of S201. Note that, when returning to the processing of S201, the contents stored in the memory can be reset. If the predetermined time has not elapsed, the vibration amplitude value analysis unit 10 advances the process to S213.
　Ｓ２１３において、第２の実施形態におけるＳ１０９の動作と同様に、信号判定部６は、連続する所定の数点ごとの傾きが、判定用閾値以上でかつ判定用継続回数を越えた値で観測された場合、Ｓ２１４に処理を進め機器の故障が予測できると判断する。 In S213, similar to the operation of S109 in the second embodiment, the signal determination unit 6 is observed with a value at which the slope of every predetermined number of consecutive points is equal to or greater than the determination threshold and exceeds the determination continuation count. If YES, the process proceeds to S214, and it is determined that a device failure can be predicted.
　しかし、Ｓ２１３において、信号判定部６は、上記の条件を満たさない場合は、特定の周波数ｆにおける振幅値の故障解析を終了する。そして、Ｓ２０１の処理に戻る。なお、Ｓ１の処理に戻る際に、メモリに保存された内容をリセットすることもできる。 However, in S213, the signal determination unit 6 ends the failure analysis of the amplitude value at the specific frequency f when the above condition is not satisfied. Then, the process returns to S201. In addition, when returning to the process of S1, the content preserve | saved at memory can also be reset.
　また、第４の実施形態における故障予測システム１の動作は、Ｓ２１２とＳ２１３は動作の順番を入れ替えることも可能である。 Also, the operation of the failure prediction system 1 in the fourth embodiment can be switched in the order of operations in S212 and S213.
　[効果の説明]第３の実施形態における効果について説明する。第３の実施形態における故障予測システム１は、残差設定部１１が前回の測定値までの所定の数点ごとの傾きから、今回の測定値の残差を算出する。そして、残差設定部１１は残差閾値を設定し、今回の測定値の残差が残差閾値を越えた場合に除外を行う。 [Explanation of Effects] Effects in the third embodiment will be described. In the failure prediction system 1 according to the third embodiment, the residual setting unit 11 calculates the residual of the current measurement value from the slopes of several predetermined points up to the previous measurement value. Then, the residual setting unit 11 sets a residual threshold, and performs exclusion when the residual of the current measurement value exceeds the residual threshold.
　第３の実施形態における故障予測システム１は、残差設定部１１が残差閾値を設定することで、雑音等に起因した振動を除外できるため相関係数が高い直線回帰が可能となる。また故障の進行に関与する振動に対して、直線回帰を行い直線の傾きから故障診断を行うことで、精度の高い故障予測を行うことができる。 In the failure prediction system 1 according to the third embodiment, since the residual setting unit 11 sets a residual threshold, vibration caused by noise or the like can be excluded, and thus linear regression with a high correlation coefficient is possible. In addition, it is possible to perform highly accurate failure prediction by performing linear regression on the vibration related to the progress of the failure and performing failure diagnosis from the slope of the straight line.
　第３の実施例における故障予測システム１では、残差設定部１１は残差閾値の範囲外である測定値が２点以上連続して計算された場合は、測定値の除外は行わない。そのため、故障の要因となる可能性が高い振動による測定値の除外を回避することができる。 In the failure prediction system 1 in the third embodiment, the residual setting unit 11 does not exclude measured values when two or more measured values that are outside the range of the residual threshold are calculated continuously. Therefore, it is possible to avoid the exclusion of the measurement value due to vibrations that are likely to cause failure.
　[第４の実施形態]本発明の第４の実施形態を、図面を用いて詳細に説明する。 [Fourth Embodiment] A fourth embodiment of the present invention will be described in detail with reference to the drawings.
　[構成の説明]第４の実施形態は、第３の実施形態と同様な図７のシステム全体構成図で構成されている。 [Description of Configuration] The fourth embodiment is configured by the entire system configuration diagram of FIG. 7 similar to the third embodiment.
　[動作の説明]次に、第４の実施形態における故障予測システム１の動作を図９Ａ、図９Ｂのフローチャートを用いて説明する。 [Description of Operation] Next, the operation of the failure prediction system 1 in the fourth embodiment will be described with reference to the flowcharts of FIGS. 9A and 9B.
　Ｓ３０１において、振動計測器３は、機器の周波数と振動振幅値とを計測し、計測した周波数と振動振幅値の信号を信号処理器７の信号変換部４へ送る。次にＳ３０２処理を進める。 In S <b> 301, the vibration measuring device 3 measures the frequency and vibration amplitude value of the device, and sends a signal of the measured frequency and vibration amplitude value to the signal conversion unit 4 of the signal processor 7. Next, S302 processing is advanced.
　Ｓ３０２において、信号変換部４と振動解析開始部９は第３の実施形態のＳ２０２における動作と同様の信号解析開始処理を行う。次に、Ｓ３０３に処理を進める。 In S302, the signal conversion unit 4 and the vibration analysis start unit 9 perform a signal analysis start process similar to the operation in S202 of the third embodiment. Next, the process proceeds to S303.
　Ｓ３０３において、振動振幅値解析部１０は異常点を除外するために測定する振幅値の下限値と上限値を設定する。次にＳ３０４に処理を進める。 In S303, the vibration amplitude value analysis unit 10 sets a lower limit value and an upper limit value of the amplitude value to be measured in order to exclude an abnormal point. Next, the process proceeds to S304.
　Ｓ３０４において、振動振幅値解析部１０は、測定値が下限値以上、上限値以下で設定される範囲以内であるか比較を行う。測定値が範囲を超えた場合はＳ３０５に処理を進める。測定値が範囲を超えない場合は、Ｓ３０６に処理を進める。 In S304, the vibration amplitude value analysis unit 10 compares whether the measured value is within a range set between the lower limit value and the upper limit value. If the measured value exceeds the range, the process proceeds to S305. If the measured value does not exceed the range, the process proceeds to S306.
　Ｓ３０５において、振動振幅値解析部１０は、上記の範囲を超えた測定を異常データとして除外する。次にＳ３０６に処理を進める。 In S305, the vibration amplitude value analysis unit 10 excludes measurement exceeding the above range as abnormal data. Next, the process proceeds to S306.
　Ｓ３０６において、振動振幅値解析部１０は、メモリ８に保存されている特定の周波数ｆにおける振幅値について時間を独立変数とした直線回帰を行う。次にＳ３０７に処理を進める。 In S306, the vibration amplitude value analysis unit 10 performs linear regression using the time as an independent variable for the amplitude value at a specific frequency f stored in the memory 8. Next, the process proceeds to S307.
　Ｓ３０７において、振動振幅値解析部１０は、連続する所定の数点ごとの傾きを算出する。次にＳ３０８に処理を進める。所定の数点とは、閾値Ａや閾値時間Ｂと同様に実際に機器が故障に至った実験のデータベースに基づいて、故障の要因となる値を設定する。 In S307, the vibration amplitude value analysis unit 10 calculates the inclination for each predetermined number of consecutive points. Next, the process proceeds to S308. As the predetermined number of points, a value that causes a failure is set on the basis of a database of experiments in which the device actually failed as in the case of the threshold A and the threshold time B.
　Ｓ３０８において、残差設定部１１は今回の測定値を含めた連続する所定の数点ごとの傾きから、次回に測定される予測値を算出する。次にＳ３０９に処理を進める。 In S <b> 308, the residual setting unit 11 calculates a predicted value to be measured next time from the slope of every predetermined number of consecutive points including the current measured value. Next, the process proceeds to S309.
　Ｓ３０９において、残差設定部１１は第３の実施形態と同様に、予測値から次回の測定における残差閾値を算出する。次にＳ３１０に処理を進める。 In S309, the residual setting unit 11 calculates a residual threshold in the next measurement from the predicted value, as in the third embodiment. Next, the process proceeds to S310.
　Ｓ３１０において、残差設定部１１は次回の測定における予測値と残差閾値とをメモリ８に保存する。 In S310, the residual setting unit 11 stores the predicted value and the residual threshold in the next measurement in the memory 8.
　Ｓ３１１において、残差設定部１１は前回の測定時にメモリ８に保存している今回の測定における予測値と、今回の測定値におけるデータから残差を算出する。次にＳ３１２に処理を進める。 In S311, the residual setting unit 11 calculates a residual from the predicted value in the current measurement stored in the memory 8 at the previous measurement and the data in the current measured value. Next, the process proceeds to S312.
　Ｓ３１２において、残差設定部１１は今回の測定値で算出された残差が、メモリ８に保存している今回の測定における残差閾値の下限値以上、上限値以下の範囲を超えた場合は、Ｓ３１３に処理を進める。今回の測定値で算出された残差が、メモリ８に保存している残差閾値の範囲を越えない場合は、Ｓ３１４に処理を進める。 In S <b> 312, the residual setting unit 11 determines that the residual calculated with the current measurement value exceeds the range between the lower limit value and the upper limit value of the residual threshold value in the current measurement stored in the memory 8. , The process proceeds to S313. If the residual calculated with the current measurement value does not exceed the residual threshold range stored in the memory 8, the process proceeds to S314.
　なお、上記において残差閾値は指定倍数とされているが、前回の測定値までの連続する所定の数点ごとの傾きに基づいて残差閾値を絶対値を変動することもできる。ただし、Ｓ３１２において残差設定部１１は、残差閾値の範囲外である測定値が２点以上連続して計測された場合は、Ｓ３１３に処理を進めない。 In the above description, the residual threshold is set to a specified multiple. However, the absolute value of the residual threshold can be changed based on the slope of every predetermined number of points up to the previous measurement value. However, in S312, the residual setting unit 11 does not advance the process to S313 when two or more measured values that are outside the range of the residual threshold are continuously measured.
　Ｓ３１３において、残差設定部１１は、残差閾値を超えた測定値を除外する。そしてＳ３１４に処理を進める。 In S313, the residual setting unit 11 excludes measurement values that exceed the residual threshold. Then, the process proceeds to S314.
　Ｓ３１４において、振動振幅値解析部１０は、所定の時間経過すると特定の周波数ｆにおける振幅値の故障解析を終了し、Ｓ３０１の処理に戻る。なお、Ｓ３０１の処理に戻る際に、メモリに保存された内容をリセットすることもできる。振動振幅値解析部１０は、所定の時間経過していなければ、Ｓ３１５に処理を進める。 In S314, the vibration amplitude value analysis unit 10 ends the failure analysis of the amplitude value at the specific frequency f when a predetermined time elapses, and returns to the process of S301. Note that the content stored in the memory can be reset when returning to the processing of S301. If the predetermined time has not elapsed, the vibration amplitude value analysis unit 10 advances the process to S315.
　Ｓ３１５において、第３の実施形態におけるＳ２１３の動作と同様に、信号判定部６は、連続する所定の数点ごとの傾きが、判定用閾値以上でかつ判定用継続回数を越えた値で観測された場合、Ｓ３１６に処理を進め機器の故障が予測できると判断する。 In S315, as in the operation of S213 in the third embodiment, the signal determination unit 6 is observed with a value at which the slope of every predetermined number of consecutive points is not less than the determination threshold and exceeds the determination continuation count. If YES in step S316, the process advances to step S316 to determine that a failure of the device can be predicted.
　しかし、Ｓ３１５において、信号判定部６は、上記の条件を満たさない場合は、特定の周波数ｆにおける振幅値の故障解析を終了する。そして、Ｓ３０１の処理に戻る。なお、Ｓ１の処理に戻る際に、メモリに保存された内容をリセットすることもできる。 However, in S315, when the above condition is not satisfied, the signal determination unit 6 ends the failure analysis of the amplitude value at the specific frequency f. Then, the process returns to S301. In addition, when returning to the process of S1, the content preserve | saved at memory can also be reset.
　また、第４の実施形態における故障予測システム１の動作はＳ３１４とＳ３１５は動作の順番を入れ替えることも可能である。 Also, the operation of the failure prediction system 1 in the fourth embodiment can be switched in order of operations in S314 and S315.
　図９Ａ、図９Ｂのフローチャートにおける、Ａで示したＳ３０８からＳ３１０の動作の工程は、故障予測システム１を動作させるＣＰＵとは、異なるＣＰＵを利用することができる。例えば、故障予測システム１がＣＰＵを複数個有している場合や、また故障予測システム１が適用される機器のＣＰＵや、故障予測の対象となる機器とは異なる外部装置が有するＣＰＵと接続することで、故障予測システム１を動作させるＣＰＵとは、異なるＣＰＵを利用することができる。 9A and 9B in the flowcharts of S308 to S310 indicated by A can use a CPU different from the CPU that operates the failure prediction system 1. For example, when the failure prediction system 1 has a plurality of CPUs, or connected to a CPU of a device to which the failure prediction system 1 is applied, or a CPU of an external device different from the device that is the target of failure prediction Thus, a CPU different from the CPU that operates the failure prediction system 1 can be used.
　[効果の説明]第４の実施形態における効果について説明する。第４の実施形態における故障予測システム１は、残差設定部１１が今回の測定値を含めた連続する所定の数点ごとの傾きから、次回に測定される予測値を算出する。そして残差設定部１１は、予測値から次回の測定における残差閾値を算出し、予測値と残差閾値を算出しメモリ８に保存する。 [Explanation of Effects] Effects in the fourth embodiment will be described. In the failure prediction system 1 according to the fourth embodiment, the residual setting unit 11 calculates a predicted value to be measured next time from the slopes of a predetermined number of consecutive points including the current measured value. Then, the residual setting unit 11 calculates a residual threshold in the next measurement from the predicted value, calculates the predicted value and the residual threshold, and stores them in the memory 8.
　そのため、第４の実施形態における故障予測システム１は、今回の測定値と、メモリ８に保存している予測値の差である残差を算出する。そして、残差とメモリ８に保存している残差閾値とを比較することで、測定値が残差閾値を越えているか判断することができる。予測値と残差閾値を算出してメモリ８に保存する手段と、残差を算出し残差閾値と比較する手段を平行して行うことで、残差閾値外の測定値である雑音等に起因した振動を除外する動作処理を迅速に行うことができる。 Therefore, the failure prediction system 1 in the fourth embodiment calculates a residual that is a difference between the current measurement value and the prediction value stored in the memory 8. Then, by comparing the residual with the residual threshold stored in the memory 8, it can be determined whether the measured value exceeds the residual threshold. By performing in parallel the means for calculating the prediction value and the residual threshold and storing it in the memory 8 and the means for calculating the residual and comparing it with the residual threshold, noise or the like that is a measurement value outside the residual threshold is obtained. It is possible to quickly perform an operation process that excludes the caused vibration.
　また、第４の実施形態における故障予測システム１は、Ｓ３０８からＳ３１０の動作における予測値と残差閾値の算出を、故障予測システム１を動作させるＣＰＵとは異なるＣＰＵで行うことで、予測値と残差閾値の算出を迅速に行うことができる。その結果、測定値が残差閾値で定められる範囲であるかの判断も迅速に行うことができる。 Moreover, the failure prediction system 1 in the fourth embodiment performs the calculation of the prediction value and the residual threshold in the operations from S308 to S310 by using a CPU different from the CPU that operates the failure prediction system 1, thereby obtaining the prediction value. The residual threshold value can be calculated quickly. As a result, it can be quickly determined whether the measured value is within the range determined by the residual threshold.
　[第５の実施形態]（故障予測システム１を内蔵した場合）本発明の第５の実施形態として、コンピュータ装置１３に搭載される記憶装置２に、本発明の故障予測システム１を適用した場合を説明する。 [Fifth Embodiment] (When the Failure Prediction System 1 is Built in) As a fifth embodiment of the present invention, the failure prediction system 1 of the present invention is applied to the storage device 2 mounted on the computer device 13. Will be explained.
　[構成の説明]図１０にシステム全体構成図を示す。故障予測システム１は、故障予測の対象となるコンピュータ装置１３の内部に位置し、振動計測器３と信号処理器７とメモリ８を備えている。 [Description of configuration] Fig. 10 shows the overall system configuration. The failure prediction system 1 is located inside a computer device 13 that is a target of failure prediction, and includes a vibration measuring device 3, a signal processor 7, and a memory 8.
　コンピュータ装置１３は、記憶装置２と制御回路１２を備えている。 The computer device 13 includes a storage device 2 and a control circuit 12.
　信号処理器７は、信号変換部４と信号解析部５と信号判定部６とを備えており、各々が第４の実施形態と同様の接続関係と機能を有している。また信号解析部５が備える、振動解析開始部９と振動振幅値解析部１０と残差設定部１１とについても第４の実施例と同様な接続関係を有している。 The signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6, each having the same connection relationship and function as in the fourth embodiment. Further, the vibration analysis starting unit 9, the vibration amplitude value analyzing unit 10, and the residual setting unit 11 included in the signal analyzing unit 5 have the same connection relationship as in the fourth embodiment.
　相違点として振動計測器３は、コンピュータ装置１３の記憶装置２に取り付けられ、信号解析部５の振動振幅値解析部１０は、メモリ８と接続している。また信号判定部６は、制御回路１２と接続しており、故障と判断すると制御回路１２に故障の情報を送信する。 As a difference, the vibration measuring instrument 3 is attached to the storage device 2 of the computer device 13, and the vibration amplitude value analyzing unit 10 of the signal analyzing unit 5 is connected to the memory 8. The signal determination unit 6 is connected to the control circuit 12 and transmits failure information to the control circuit 12 when it is determined that there is a failure.
　[動作の説明]次に動作について説明する。 [Description of operation] Next, the operation will be described.
　動作については、第１の実施形態から第４の実施形態のいずれかの動作で動く。 The operation is the same as that of the first embodiment to the fourth embodiment.
　相違点として、信号判定部６は故障と予測するとコンピュータ装置１３の制御回路１２に故障の情報を送信する。 As a difference, if the signal determination unit 6 predicts a failure, the signal determination unit 6 transmits failure information to the control circuit 12 of the computer device 13.
　また第５の実施形態は故障予測の動作を複数回行うこともできる。信号判定部６が故障を予測できると判断した後、制御回路１２は故障を予測できると判断した回数をカウントし、動作が最初に戻る。 In the fifth embodiment, the failure prediction operation can be performed a plurality of times. After the signal determination unit 6 determines that the failure can be predicted, the control circuit 12 counts the number of times that the failure is predicted and the operation returns to the beginning.
　制御回路１２は、故障を予測できると判断した回数が１回目の場合はコンピュータ装置１３の利用者には警告の情報を示す。また制御回路１２は、故障を予測できると判断した回数が２度目の場合は、インターネットや配線などを通して他の装置にデータを移動し、３度目の場合は故障が起きる可能性が高いとしてコンピュータ装置１３の電源を強制終了する。 When the number of times that the control circuit 12 determines that a failure can be predicted is the first time, the control circuit 12 indicates warning information to the user of the computer device 13. The control circuit 12 moves the data to another device through the Internet or wiring if the number of times it is determined that the failure can be predicted is the second time, and if it is the third time, the computer device is assumed to be likely to fail. 13 is forcibly terminated.
　[効果の説明]第５の実施形態における効果について説明する。第５の実施形態における故障予測システム１は、第４の実施形態までの効果と同様に、コンピュータ装置１３の通常動作の性能を低下させず、かつ故障予測精度が高い故障予測システム１を提供することができる。 [Description of Effects] Effects in the fifth embodiment will be described. The failure prediction system 1 in the fifth embodiment provides a failure prediction system 1 that does not deteriorate the performance of the normal operation of the computer device 13 and has high failure prediction accuracy, similarly to the effects up to the fourth embodiment. be able to.
　また図１１のように故障予測システム１は、コンピュータ装置１３に内蔵することで故障予測の解析によるデータの保存をコンピュータ装置１３の装置メモリ１４を利用することができる。そのため故障予測システム１は、新たにメモリを設ける必要はなく低コストでコンパクトな構成で上述の効果を達成することができる。 Further, as shown in FIG. 11, the failure prediction system 1 can be stored in the computer device 13 to use the device memory 14 of the computer device 13 to store data based on failure prediction analysis. Therefore, the failure prediction system 1 does not need to provide a new memory, and can achieve the above-described effects with a low-cost and compact configuration.
　図１２のように故障予測システム１は、外部装置１５に外部メモリ１６を設けることもできる。利用者は故障予測システム１が導入してあるコンピュータ装置１３を使用した後に、扱うデータ量に応じて必要な容量の外部メモリ１６を設定することができる。そのため利用者は、無駄なコストをかける必要がなく、利用状況に応じて故障予測システム１の外部メモリ１６を選択することができる。 As shown in FIG. 12, the failure prediction system 1 can also provide an external memory 16 in the external device 15. The user can set the external memory 16 having a necessary capacity according to the amount of data to be handled after using the computer device 13 in which the failure prediction system 1 has been introduced. Therefore, the user does not need to spend unnecessary costs, and can select the external memory 16 of the failure prediction system 1 according to the usage situation.
　また、制御回路１２が故障と予測できると判断した回数をカウントして、そのカウント数に応じて、利用者に警告の情報の表示や、他の装置へのデータの移動や、コンピュータ装置１３の強制終了を自動的に行う。そのため、利用者が警告の情報に気付かない場合に、故障と予測できると判断した回数がカウントされ危険度が増すと、データの移動やコンピュータ装置１３の強制終了などの処理を自動に行うことで、データの安全性が確保される。 Further, the number of times that the control circuit 12 determines that the failure can be predicted is counted, and according to the counted number, warning information is displayed to the user, data is transferred to another device, and the computer device 13 Forced termination is performed automatically. For this reason, when the user is not aware of the warning information, the number of times it is determined that a failure can be predicted is counted, and when the risk increases, processing such as data movement or forced termination of the computer device 13 is automatically performed. Data safety is ensured.
　[第６の実施形態]（故障予測システム１を外付けした場合）[構成の説明]図１３にシステム全体構成図を示す。故障予測システム１は、故障予測の対象となるコンピュータ装置１３の外部に位置し、コンピュータ装置１３と外部から接続している。故障予測システム１は振動計測器３と信号処理器７とメモリ８を備えている。 [Sixth Embodiment] (When the failure prediction system 1 is externally attached) [Description of Configuration] FIG. The failure prediction system 1 is located outside the computer device 13 that is a target of failure prediction, and is connected to the computer device 13 from the outside. The failure prediction system 1 includes a vibration measuring instrument 3, a signal processor 7, and a memory 8.
　コンピュータ装置１３は、記憶装置２とメモリ８と制御回路１２を備えている。 The computer device 13 includes a storage device 2, a memory 8, and a control circuit 12.
　信号処理器７は、信号変換部４と信号解析部５と信号判定部６を備えており、各々が第５の実施形態と同様の接続関係と機能を有している。また信号解析部５が備える、振動解析開始部９と振動振幅値解析部１０と残差設定部１１とについても第５の実施例と同様な接続関係を有している。 The signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6, each having the same connection relationship and function as in the fifth embodiment. Further, the vibration analysis start unit 9, the vibration amplitude value analysis unit 10, and the residual setting unit 11 included in the signal analysis unit 5 have the same connection relationship as in the fifth embodiment.
　また故障予測システム１の動作については、第５の実施形態と同様であるので省略する。 Also, the operation of the failure prediction system 1 is the same as that of the fifth embodiment, and therefore will be omitted.
　[効果の説明]第６の実施形態における効果について説明する。第６の実施形態における故障予測システム１は、第５の実施形態の効果と同様に、コンピュータ装置１３の通常動作の性能を低下させず、かつ故障予測精度が高い故障予測システム１を提供できる。 [Explanation of Effects] Effects in the sixth embodiment will be described. The failure prediction system 1 in the sixth embodiment can provide the failure prediction system 1 with high failure prediction accuracy without degrading the performance of the normal operation of the computer device 13, similarly to the effect of the fifth embodiment.
　また故障予測システム１は、コンピュータ装置１３に外部から接続している。そのため故障予測システム１を必要としない利用者は、故障予測システム１をコンピュータ装置１３に最初から搭載することによる価格の上昇を抑えることができる。そして利用者は、コンピュータ装置１３を利用後に、利用者の必要に応じて故障予測システム１を導入することができる。 The failure prediction system 1 is connected to the computer device 13 from the outside. Therefore, a user who does not need the failure prediction system 1 can suppress an increase in price caused by mounting the failure prediction system 1 on the computer device 13 from the beginning. Then, after using the computer device 13, the user can introduce the failure prediction system 1 as required by the user.
　また図１５のように故障予測システム１は、コンピュータ装置１３に内蔵している装置メモリ１４を使用することができる。そのため利用者は新たにメモリを必要とすることがないので、低コストでコンパクトな構成で上述の効果を達成することができる。 Further, as shown in FIG. 15, the failure prediction system 1 can use the device memory 14 built in the computer device 13. Therefore, since the user does not need a new memory, the above-described effects can be achieved with a low-cost and compact configuration.
　図１６のように故障予測システム１は、外部装置１５に外部メモリ１６を設けることもできる。利用者は、コンピュータ装置１３に故障予測システム１を導入して使用した後に、利用者の扱うデータ量に応じて必要な容量の外部メモリ１６を設定することができる。そのため、利用者は無駄なコストをかける必要がなく、利用状況に応じて故障予測システム１の外部メモリ１６を選択することができる。 As shown in FIG. 16, the failure prediction system 1 can also provide an external memory 16 in the external device 15. The user can set the external memory 16 having a necessary capacity according to the amount of data handled by the user after introducing and using the failure prediction system 1 in the computer device 13. Therefore, the user does not need to spend unnecessary cost, and can select the external memory 16 of the failure prediction system 1 according to the usage situation.
　また、第５の実施形態と同様に制御回路１２が故障と予測できると判断した回数をカウントして、そのカウント数に応じて、利用者に警告の情報の表示や、他の装置へのデータの移動や、コンピュータ装置１３の強制終了を自動的に行う。そのため、利用者が警告の情報に気付かない場合などに、故障と予測できると判断した回数がカウントされ危険度が増すと、データの移動やコンピュータ装置１３の強制終了などの処理を自動に行うことで、データの安全性が確保される。 Similarly to the fifth embodiment, the control circuit 12 counts the number of times it is determined that a failure can be predicted, and displays warning information to the user or data to other devices according to the counted number. And the forced termination of the computer device 13 are automatically performed. For this reason, when the user is not aware of the warning information, the number of times that it is determined that the failure can be predicted is counted, and when the risk increases, processing such as data movement or forced termination of the computer device 13 is automatically performed. As a result, the safety of the data is ensured.
　[第７の実施形態]（故障予測システム１をハードディスク装置１７に適用した場合）本発明の第７の実施形態として、本発明の故障予測システム１を組み込んだハードディスク装置１７を説明する。 [Seventh Embodiment] (When Failure Prediction System 1 is Applied to Hard Disk Device 17) A hard disk device 17 incorporating the failure prediction system 1 of the present invention will be described as a seventh embodiment of the present invention.
　[構成の説明]図１６にシステム全体構成図を示す。故障予測システム１は、故障予測の対象となるハードディスク装置１７の内部に位置し、振動計測器３と信号処理器７とメモリ８を備えている。 [Description of configuration] Fig. 16 shows the overall system configuration. The failure prediction system 1 is located inside a hard disk device 17 that is a target of failure prediction, and includes a vibration measuring device 3, a signal processor 7, and a memory 8.
　ハードディスク装置１７は、ハードディスク制御回路１８を備えている。 The hard disk device 17 includes a hard disk control circuit 18.
　信号処理器７は、信号変換部４と信号解析部５と信号判定部６を備えており、各々が第６の実施形態と同様の接続関係と機能を有している。また信号解析部５が備える、振動解析開始部９と振動振幅値解析部１０と残差設定部１１とについても第６の実施例と同様な接続関係を有している。 The signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, and a signal determination unit 6, each having the same connection relationship and function as in the sixth embodiment. Further, the vibration analysis starting unit 9, the vibration amplitude value analyzing unit 10, and the residual setting unit 11 included in the signal analyzing unit 5 have the same connection relationship as in the sixth embodiment.
　相違点として振動計測器３は、ハードディスク装置１７に取り付けられ、信号判定部６はハードディスク制御回路１８と接続している。信号判定部６は、故障と判断するとハードディスク制御回路１８に故障の情報を送信する。 As a difference, the vibration measuring instrument 3 is attached to the hard disk device 17, and the signal determination unit 6 is connected to the hard disk control circuit 18. When the signal determination unit 6 determines that a failure has occurred, the signal determination unit 6 transmits failure information to the hard disk control circuit 18.
　[動作の説明]次に動作について説明する。 [Description of operation] Next, the operation will be described.
　また故障予測システム１の動作については、第６の実施形態と同様であるので省略する。 Also, the operation of the failure prediction system 1 is the same as that in the sixth embodiment, and therefore will be omitted.
　相違点として振動計測器３は、ハードディスク装置１７の振動を計測する。また信号判定部６は、故障と予測できると判断するとハードディスク装置１７のハードディスク制御回路１８に故障の情報を送信する。 As a difference, the vibration measuring instrument 3 measures the vibration of the hard disk device 17. If the signal determination unit 6 determines that a failure can be predicted, the signal determination unit 6 transmits failure information to the hard disk control circuit 18 of the hard disk device 17.
　また第７の実施形態は、第６の実施形態と同様に故障予測の動作を複数回行うこともできる。信号判定部６が故障を予測できると判断した後、ハードディスク制御回路１８は故障を予測できると判断した回数をカウントし、動作が最初に戻る。 Also, in the seventh embodiment, the failure prediction operation can be performed a plurality of times as in the sixth embodiment. After the signal determination unit 6 determines that the failure can be predicted, the hard disk control circuit 18 counts the number of times that the failure is predicted and the operation returns to the beginning.
　ハードディスク制御回路１８は、故障を予測できると判断した回数に応じてハードディスク装置１７が搭載された機器に警告の情報を送信する。 The hard disk control circuit 18 transmits warning information to a device on which the hard disk device 17 is mounted according to the number of times it is determined that a failure can be predicted.
　[効果の説明]第７の実施形態における効果について説明する。第７の実施形態における故障予測システム１は、ハードディスク装置１７に組み込まれることによって特定のコンピュータ装置１３などに限らず、ハードディスク装置１７が搭載される様々な機器に故障予測システム１を適用することができる。 [Description of Effects] Effects in the seventh embodiment will be described. The failure prediction system 1 in the seventh embodiment is not limited to a specific computer device 13 by being incorporated in the hard disk device 17, and the failure prediction system 1 can be applied to various devices in which the hard disk device 17 is mounted. it can.
　また図１７のような故障予測システム１は、ハードディスク装置１７に内蔵することで故障予測の解析によるデータの保存をハードディスク装置１７の装置メモリ１４を利用することができる。そのため故障予測システム１は、新たにメモリを設ける必要はなく低コストでコンパクトな構成で上述の効果を達成することができる。 Further, the failure prediction system 1 as shown in FIG. 17 can be stored in the hard disk device 17 to use the device memory 14 of the hard disk device 17 to store data by analyzing the failure prediction. Therefore, the failure prediction system 1 does not need to provide a new memory, and can achieve the above-described effects with a low-cost and compact configuration.
　ハード制御回路１２は故障と予測できると判断した回数をカウントし、そのカウント数に応じて、ハードディスク装置１７が搭載された機器に警告の情報を送信する。そのため機器は、送信された情報の回数に応じて危険度を認識することができ、データの移動などの安全性を確保する処理を行うことが出来る。 The hardware control circuit 12 counts the number of times it is determined that a failure can be predicted, and transmits warning information to a device on which the hard disk device 17 is mounted according to the count. Therefore, the device can recognize the degree of danger according to the number of transmitted information, and can perform processing for ensuring safety such as data movement.
　本発明の実施例として、コンピュータ装置１３に搭載される記憶装置２に、本発明の故障予測システム１を適用した場合を説明する。 As an example of the present invention, a case where the failure prediction system 1 of the present invention is applied to the storage device 2 mounted on the computer device 13 will be described.
　図１０のシステム全体構成図において、実施を行った。故障予測システム１は、故障予測の対象となるコンピュータ装置１３の内部に位置し、振動計測器３と信号処理器７とを備えている。 Executed in the overall system configuration diagram of FIG. The failure prediction system 1 is located inside a computer device 13 that is a target of failure prediction, and includes a vibration measuring device 3 and a signal processor 7.
　コンピュータ装置１３は記憶装置２と制御回路１２を備えている。 The computer device 13 includes a storage device 2 and a control circuit 12.
　信号処理器７は信号変換部４と信号解析部５と信号判定部６とメモリ８とを備えている。 The signal processor 7 includes a signal conversion unit 4, a signal analysis unit 5, a signal determination unit 6, and a memory 8.
　振動計測器３は、コンピュータ装置１３の記憶装置２に取り付けられ、故障予測の対象となる機器の振動を計測する機能を有している。 The vibration measuring instrument 3 is attached to the storage device 2 of the computer device 13 and has a function of measuring the vibration of a device that is a target of failure prediction.
　信号変換部４は、振動計測器３と信号解析部５に接続され、振動のアナログ信号をデジタル信号へ変換する機能を有している。 The signal converter 4 is connected to the vibration measuring instrument 3 and the signal analyzer 5 and has a function of converting an analog signal of vibration into a digital signal.
　信号解析部５は、振動解析開始部９と振動振幅値解析部１０と残差設定部１１とを備えている。 The signal analysis unit 5 includes a vibration analysis start unit 9, a vibration amplitude value analysis unit 10, and a residual setting unit 11.
　振動解析開始部９は、信号変換部４と接続しており、振動のデジタル信号に対して、振幅値の大きさを周波数の値から見た振動スペクトルの計算を行う。そして特定の周波数ｆにおいて、振動の振幅値が所定の閾値を越え、かつ所定の時間以上継続すると、振幅値について解析を開始する。 The vibration analysis start unit 9 is connected to the signal conversion unit 4 and calculates the vibration spectrum of the vibration digital signal as seen from the frequency value of the amplitude value. When the vibration amplitude value exceeds a predetermined threshold value and continues for a predetermined time or more at a specific frequency f, analysis of the amplitude value is started.
　振動振幅値解析部１０は、振動解析開始部９と接続しており、振動の振幅値について時間を独立変数とした直線回帰を行い、各計測時間ごとに直線の傾きを計算する。 The vibration amplitude value analysis unit 10 is connected to the vibration analysis start unit 9, and performs linear regression with the time as an independent variable for the vibration amplitude value, and calculates the slope of the line for each measurement time.
　残差設定部１１は、振動振幅値解析部１０と接続しており、各計測時間ごとの傾きから残差閾値を計算し、残差閾値を越えた残差を除外する機能を有する。 The residual setting unit 11 is connected to the vibration amplitude value analyzing unit 10 and has a function of calculating a residual threshold value from the slope for each measurement time and excluding the residual exceeding the residual threshold value.
　信号判定部６は、信号解析部５とコンピュータ装置１３の制御回路１２と接続しており、故障と判断すると制御回路１２に故障の情報を送信する。 The signal determination unit 6 is connected to the signal analysis unit 5 and the control circuit 12 of the computer device 13, and transmits failure information to the control circuit 12 when it is determined that there is a failure.
　本実施例では、振動計測器３は振動センサ１９を用いた。長さが７ｍｍ、幅が７ｍｍ、高さが５ｍｍの圧電型加速度センサを、記憶装置２の表面中心に設置した。 In the present embodiment, the vibration measuring instrument 3 uses the vibration sensor 19. A piezoelectric acceleration sensor having a length of 7 mm, a width of 7 mm, and a height of 5 mm was installed at the center of the surface of the storage device 2.
　また振動計測器３は、振動センサ１９の代わりに音響マイクロホン２０を用いて記憶装置２の振動を計測しても同様の効果を得ることができる。 Further, the vibration measuring device 3 can obtain the same effect even if the vibration of the storage device 2 is measured using the acoustic microphone 20 instead of the vibration sensor 19.
　本実施例で用いた振動センサ１９の自己共振周波数は５０ｋＨｚ以上であり、測定する周波数帯域は、１Ｈｚから２０ｋＨｚとした。計測を開始する振動スペクトルの振幅値の閾値Aは０．５μｍ、閾値Ａを継続するする時間の閾値Ｂを１０秒に設定した。振幅データは５秒間隔で計測し、また直線回帰による傾きは連続する１０点ごとに算出し、残差閾値は±５倍とした。 The self-resonant frequency of the vibration sensor 19 used in this example is 50 kHz or more, and the frequency band to be measured is 1 Hz to 20 kHz. The threshold value A of the amplitude value of the vibration spectrum to start measurement was set to 0.5 μm, and the threshold value B of the time for continuing the threshold value A was set to 10 seconds. The amplitude data was measured at intervals of 5 seconds, and the slope by linear regression was calculated every 10 consecutive points, and the residual threshold was set to ± 5 times.
　判定条件は、判定用閾値を直線回帰における傾きが初期の傾きに対して１０倍に設定し、判定用継続回数を５回とした。振動センサ１９の方式、寸法、設置箇所、測定周波数帯域、自己共振周波数、計測開始条件および判定条件は、これに限定されるものでない。 The judgment conditions were such that the judgment threshold was set to 10 times the slope in linear regression with respect to the initial slope, and the number of judgment continuations was five. The method, dimensions, installation location, measurement frequency band, self-resonant frequency, measurement start condition and determination condition of the vibration sensor 19 are not limited to this.
　次に、実施例における動作について説明する。 Next, the operation in the embodiment will be described.
　振動計測器３は、振動センサ１９により記憶装置２の表面中心における振動を計測し、増幅した振動のアナログ信号を信号処理器７の信号変換部４へ送信した。 The vibration measuring instrument 3 measured the vibration at the center of the surface of the storage device 2 with the vibration sensor 19 and transmitted the amplified analog signal of the vibration to the signal conversion unit 4 of the signal processor 7.
　信号変換部４において計測した振動のアナログ信号をデジタル信号へ変換し、信号解析部５に送信した。 The analog signal of vibration measured in the signal conversion unit 4 was converted into a digital signal and transmitted to the signal analysis unit 5.
　信号解析部５における振動解析開始部９は信号変換部４からの振動のデジタル信号に対して離散型の高速フーリエ変換処理を適用して、振動スペクトルの振幅値により解析を行った。解析の周波数範囲は１Ｈｚから２０ＫＨｚとした。その結果、周波数が２００Hzにおいて、振動スペクトルの振幅値が閾値Ａである０．５μｍを超えた値が観測されため、振動解析開始部９はデータをメモリ８に保存した。 The vibration analysis start unit 9 in the signal analysis unit 5 applied a discrete fast Fourier transform process to the vibration digital signal from the signal conversion unit 4 and analyzed the amplitude value of the vibration spectrum. The frequency range of analysis was 1 Hz to 20 KHz. As a result, at a frequency of 200 Hz, a value in which the amplitude value of the vibration spectrum exceeded the threshold value A of 0.5 μm was observed, so the vibration analysis starting unit 9 saved the data in the memory 8.
　次に振動解析開始部９は、２００Hzにおいて振動の振幅値が所定の閾値Ａである０．５μｍを超えたデータが、閾値時間Ｂ以上である１０秒以上観測されメモリに保存された。 Next, the vibration analysis starting unit 9 observed that the amplitude value of vibration exceeded a predetermined threshold A of 0.5 μm at 200 Hz was observed for 10 seconds or more which is the threshold time B or longer and stored in the memory.
　そこで、振動振幅値解析部１０は２００Hzにおける振幅値の解析を開始し、２００Ｈｚにおける振幅値の時間変化のデータをメモリ８に保存した。 Therefore, the vibration amplitude value analysis unit 10 started to analyze the amplitude value at 200 Hz, and stored the time change data of the amplitude value at 200 Hz in the memory 8.
　振動振幅値解析部１０はメモリ８に保存された２００Hzにおける振幅値について時間を独立変数とした直線回帰を行い、連続する１０点ごとの直線の傾きを算出した。 The vibration amplitude value analysis unit 10 performed linear regression using time as an independent variable for the amplitude value at 200 Hz stored in the memory 8, and calculated the slope of the straight line every 10 consecutive points.
　残差設定部１１は各計測時間における残差を算出した。残差設定部１１は、直近の連続する１０点における直線回帰による傾きから次に測定される予測値を算出し、傾きを±５倍して算出した予測値の範囲を残差閾値とした。そして残差設定部１１は、予測値と残差閾値のデータをメモリ８に保存した。 The residual setting unit 11 calculated the residual at each measurement time. The residual setting unit 11 calculates a predicted value to be measured next from the slope obtained by linear regression at the last 10 consecutive points, and sets the range of the predicted value calculated by multiplying the slope by ± 5 as a residual threshold. The residual setting unit 11 stores the prediction value and residual threshold data in the memory 8.
　残差設定部１１は、測定値とメモリ８に保存された予測値から残差を算出する。残差設定部１１は、残差が残差閾値の範囲外である測定値は除外した。 The residual setting unit 11 calculates a residual from the measured value and the predicted value stored in the memory 8. The residual setting unit 11 excludes measurement values whose residuals are outside the residual threshold range.
　ただし、残差閾値の範囲外である測定値が２点以上連続して計測された場合は、２点とも除外は行わかった。 However, if two or more measured values that are outside the range of the residual threshold were measured continuously, both points were not excluded.
　これにより雑音等に起因した振動値、一過性の振動、突発的に起こる環境振動、計測ミスなどを除外することで相関係数が高い直線回帰が可能となり故障予測精度を高めることができる。 This eliminates vibration values due to noise, transient vibrations, sudden environmental vibrations, measurement errors, etc., thereby enabling linear regression with a high correlation coefficient and improving failure prediction accuracy.
　判定用閾値は、振幅値の計測開始後、最初に算出した１０点における直線回帰の傾きに対して１０倍に設定した。 The threshold value for determination was set to 10 times the slope of linear regression at the first 10 points calculated after the start of amplitude value measurement.
　信号判定部６は測定値の連続した１０点における傾きの値が、判定用閾値をこえているかどうか判定した。 The signal determination unit 6 determines whether the value of the slope at 10 consecutive measurement values exceeds the determination threshold.
　信号判定部６は、２００Ｈｚにおける振幅値の時間変化の測定を開始後１００秒後に、連続した１０点の傾きが判定用閾値を超えた状態で、５回以上連続して観測された。そして信号判定部６は、記憶装置２の故障が予測されると判断し、コンピュータ装置１３の制御回路１２へ故障の情報を送信した。 The signal determination unit 6 was continuously observed five times or more in a state in which the slope of 10 consecutive points exceeded the determination threshold, 100 seconds after starting the measurement of the time variation of the amplitude value at 200 Hz. The signal determination unit 6 determines that a failure of the storage device 2 is predicted, and transmits failure information to the control circuit 12 of the computer device 13.
　その後、評価した記憶装置を意図的に連続動作させたところ、計測開始より３５分後にハードディスクは読み書きできず故障となった。今回の実施例における実験結果と、他の故障要因に本発明の故障予測システムを適用した実験結果を図１８に示す。 After that, when the evaluated storage device was intentionally operated continuously, the hard disk could not be read and written after 35 minutes from the start of measurement, resulting in a failure. FIG. 18 shows experimental results in this example and experimental results in which the failure prediction system of the present invention is applied to other failure factors.
　本実施例では、故障に起因した特定の周波数ｆの振動の振幅を直接抽出および計測し、その振幅値の推移は残差を利用して回帰直線の精度を高めている。したがって、記憶装置２の故障予測を高い精度で実現できる。 In this embodiment, the vibration amplitude of a specific frequency f caused by a failure is directly extracted and measured, and the transition of the amplitude value uses the residual to improve the accuracy of the regression line. Therefore, failure prediction of the storage device 2 can be realized with high accuracy.
　以上、実施の形態を参照して本願発明を説明したが、本願発明は上記によって限定されるものではない。本願発明の構成や詳細には、発明のスコープ内で当業者が理解し得る様々な変更をすることができる。 The present invention has been described above with reference to the embodiment, but the present invention is not limited to the above. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the invention.
　この出願は、２００９年３月２７日に出願された日本出願特願２００９－０７９０７９を基礎とする優先権を主張し、その開示の全てをここに取り込む。 This application claims priority based on Japanese Patent Application No. 2009-0799079 filed on Mar. 27, 2009, the entire disclosure of which is incorporated herein.
１：故障予測システム
２：記憶装置
３：振動計測器
４：信号変換部
５：信号解析部
６：信号判定部
７：信号処理器
８： メモリ
９：振動解析開始部
１０：振動振幅値解析部
１１：残差設定部
１２：制御回路
１３：コンピュータ装置
１４：装置メモリ
１５：外部装置
１６：外部メモリ
１７：ハードディスク装置
１８：ハードディスク制御回路
１９：振動センサ
２０：音響マイク 1: failure prediction system 2: storage device 3: vibration measuring instrument 4: signal conversion unit 5: signal analysis unit 6: signal determination unit 7: signal processor 8: memory 9: vibration analysis start unit 10: vibration amplitude value analysis unit 11: Residual setting unit 12: Control circuit 13: Computer device 14: Device memory 15: External device 16: External memory 17: Hard disk device 18: Hard disk control circuit 19: Vibration sensor 20: Acoustic microphone

Claims (26)

  1. 　機器から発生する振動を計測する振動計測器と、
    　前記振動計測器が計測した前記振動の振幅値が所定の閾値を越えた状態で所定の時間以上継続すると、故障予測を行う信号処理器と、
    を有していることを特徴とする故障予測システム。     A vibration measuring instrument for measuring vibrations generated from the device;
    When the amplitude value of the vibration measured by the vibration measuring instrument exceeds a predetermined threshold and continues for a predetermined time, a signal processor that performs failure prediction;
    The failure prediction system characterized by having.
  2. 　前記振動計測器は、機器から発生する振動の振幅値を周波数ごとに計測することを特徴とする請求項１に記載の故障予測システム。 The failure prediction system according to claim 1, wherein the vibration measuring instrument measures an amplitude value of vibration generated from the device for each frequency.
  3. 　前記信号処理器は、前記振幅値の時間変化に基づいて故障予測を行うことを特徴とする請求項１または２に記載の故障予測システム。 The failure prediction system according to claim 1 or 2, wherein the signal processor performs failure prediction based on a time change of the amplitude value.
  4. 　前記信号処理器は、前記振幅値について時間を独立変数とした直線回帰を行い、所定の計測時間ごとの直線の傾きを算出し、前記直線の傾きの推移に基づき前記機器の故障を予測することを特徴とする請求項３に記載の故障予測システム。 The signal processor performs linear regression using time as an independent variable for the amplitude value, calculates a slope of a straight line for each predetermined measurement time, and predicts a failure of the device based on a transition of the straight line slope. The failure prediction system according to claim 3.
  5. 　前記振幅値について下限値と上限値とを設定されており、前記信号処理器は前記下限値以上、前記上限値以下の範囲を超えた前記振幅値を除外することを特徴とする請求項４に記載の故障予測システム。 5. The lower limit value and the upper limit value are set for the amplitude value, and the signal processor excludes the amplitude value that exceeds the range from the lower limit value to the upper limit value. The failure prediction system described.
  6. 　前記信号処理器は、
    　前記傾きに基づいて所定の計測時間における前記振幅値の予測値と残差閾値を算出し、
    　前記所定の計測時間における前記振幅値の測定値と前記予測値との差である残差を算出し、
    　前記残差が前記残差閾値における下限値以上、上限値以下の範囲を越えた場合に前記測定値を除外することを特徴とする請求項４または５に記載の故障予測システム。     The signal processor is
    Based on the slope, a predicted value of the amplitude value and a residual threshold value at a predetermined measurement time are calculated,
    Calculating a residual that is a difference between the measured value of the amplitude value and the predicted value at the predetermined measurement time;
    The failure prediction system according to claim 4 or 5, wherein the measured value is excluded when the residual exceeds a range between a lower limit value and an upper limit value in the residual threshold.
  7. 　前記残差閾値は、前記傾きに基づき算出することを特徴とする請求項６に記載の故障予測システム。 The failure prediction system according to claim 6, wherein the residual threshold is calculated based on the slope.
  8. 　前記信号処理器は、所定の計測時間における前記振幅値のデータを記憶するメモリを備えていることを特徴とする請求項１乃至５のいずれか１項記載の故障予測システム。 6. The failure prediction system according to claim 1, wherein the signal processor includes a memory for storing data of the amplitude value at a predetermined measurement time.
  9. 　請求項１乃至７に記載の前記故障予測システムを内部に備えていることを特徴とする電子機器。 An electronic apparatus comprising the failure prediction system according to claim 1 inside.
  10. 　前記故障予測システムは、メモリを備え、
    　前記メモリは、所定の計測時間における前記振幅値のデータを記憶することを特徴とする請求項９に記載の電子機器。     The failure prediction system includes a memory,
    The electronic device according to claim 9, wherein the memory stores data of the amplitude value at a predetermined measurement time.
  11. 　内部に故障予測システムと接続される装置メモリを備え、
    　前記装置メモリは、所定の計測時間における前記振幅値のデータを記憶することを特徴とする請求項９に記載の電子機器。     It has a device memory connected to the failure prediction system inside,
    The electronic device according to claim 9, wherein the device memory stores data of the amplitude value at a predetermined measurement time.
  12. 　前記故障予測システムは、外部装置における外部メモリと接続され、
    　前記第３のメモリは、所定の計測時間における前記振幅値のデータを記憶することを特徴とする請求項９に記載の電子機器。     The failure prediction system is connected to an external memory in an external device,
    The electronic device according to claim 9, wherein the third memory stores data of the amplitude value at a predetermined measurement time.
  13. 　請求項１乃至７に記載の前記故障予測システムと外部で接続していることを特徴とする電子機器。 An electronic device connected externally to the failure prediction system according to claim 1.
  14. 　前記故障予測システムは、メモリを備え、
    　前記メモリは、所定の計測時間における前記振幅値のデータを記憶することを特徴とする請求項１３に記載の電子機器。     The failure prediction system includes a memory,
    The electronic device according to claim 13, wherein the memory stores data of the amplitude value at a predetermined measurement time.
  15. 　内部に故障予測システムと接続される装置メモリを備え、
    　前記装置メモリは、所定の計測時間における前記振幅値のデータを記憶することを特徴とする請求項１３に記載の電子機器。     It has a device memory connected to the failure prediction system inside,
    The electronic device according to claim 13, wherein the device memory stores data of the amplitude value at a predetermined measurement time.
  16. 　前記故障予測システムは、外部装置における外部メモリと接続され、
    　前記外部メモリは、所定の計測時間における前記振幅値のデータを記憶することを特徴とする請求項１３に記載の電子機器。     The failure prediction system is connected to an external memory in an external device,
    The electronic device according to claim 13, wherein the external memory stores data of the amplitude value at a predetermined measurement time.
  17. 　前記故障予測システムが故障を予測する動作を複数回行い、故障を予測できると判断した回数に応じてデータの安全性を確保する処理を行うことを特徴とする請求項８乃至１５のいずれか１項記載の電子機器。 The operation for predicting a failure is performed a plurality of times by the failure prediction system, and processing for ensuring data safety is performed according to the number of times it is determined that a failure can be predicted. Electronic equipment described in the section.
  18. 　前記安全性を確保する処理は、警告情報を表示することを特徴とする請求項１７に記載の電子機器。 18. The electronic device according to claim 17, wherein the process for ensuring safety displays warning information.
  19. 　前記安全性を確保する処理は、データを外部の装置に移動させることを特徴とする請求項１７に記載の電子機器。 18. The electronic apparatus according to claim 17, wherein the process of ensuring the safety moves data to an external device.
  20. 　前記安全性を確保する処理は、強制終了を行うことを特徴とする請求項１７に記載の電子機器。 18. The electronic apparatus according to claim 17, wherein the process for ensuring the safety is forcibly terminated.
  21. 　機器から発生する振動を計測する第１の工程と、
    　前記振動の振幅値が所定の閾値を越え、かつ前記振幅値が所定の閾値を越えた状態で所定の時間以上継続すると、故障予測を行う第２の工程と、
    を備えていることを特徴とする故障予測方法。     A first step of measuring vibrations generated from the device;
    A second step of performing failure prediction when the amplitude value of the vibration exceeds a predetermined threshold value and continues for a predetermined time in a state where the amplitude value exceeds the predetermined threshold value;
    A failure prediction method comprising:
  22. 　前記第２の工程は、前記振幅値の時間変化に基づいて故障予測を行うことを特徴とする請求項２１に記載の故障予測方法。 The failure prediction method according to claim 21, wherein in the second step, failure prediction is performed based on a time change of the amplitude value.
  23. 　前記第２の工程は、前記振幅値について時間を独立変数とした直線回帰を行い、所定の計測時間ごとの直線の傾きを算出する工程と、
    　前記直線の傾きの推移に基づき前記機器の故障を予測する工程と、
    を備えていることを特徴とする請求項に記載２２の故障予測方法。     The second step performs a linear regression with time as an independent variable for the amplitude value, and calculates a slope of the straight line for each predetermined measurement time;
    Predicting a failure of the device based on the slope of the straight line;
    The failure prediction method according to claim 22, further comprising:
  24. 　前記第２の工程は、前記振幅値について下限値と上限値とを設定し、下限値以上、上限値以下の範囲を超えた前記振幅値を除外する工程を備えていることを特徴とする請求項２３に記載の故障予測方法。 The second step includes a step of setting a lower limit value and an upper limit value for the amplitude value, and excluding the amplitude value exceeding a range not less than the lower limit value and not more than the upper limit value. Item 24. The failure prediction method according to Item 23.
  25. 　前記第２の工程は、
    　前記傾きに基づいて所定の計測時間における前記振幅値の予測値と残差閾値とを算出する工程と、
    　前記所定の計測時間における前記振幅値の測定値と前記予測値との差である残差を算出する工程と
    　前記残差が前記残差閾値における下限値以上、上限値以下の範囲を越えた場合に前記測定値を除外する工程と、
    を備えていることを特徴とする請求項２３または２４に記載の故障予測方法。     The second step includes
    Calculating a predicted value and a residual threshold of the amplitude value at a predetermined measurement time based on the slope;
    A step of calculating a residual which is a difference between the measured value of the amplitude value and the predicted value at the predetermined measurement time; and the residual exceeds a range not less than a lower limit value and not more than an upper limit value in the residual threshold. Excluding the measured value in
    The failure prediction method according to claim 23 or 24, comprising:
  26. 　前記第２の工程は、
    　前記傾きに基づき所定の計測時間における前記振幅値の予測値と残差閾値とを算出する工程と、
    　前記予測値と前記残差閾値とをメモリに保存する工程と、
    　前記所定の計測時間における前記振幅値の測定値と前記予測値との差である残差を算出する工程と、
    　前記残差が前記残差閾値における下限値以上、上限値以下の範囲を越えた場合に前記測定値を除外することを特徴とする請求項２３または２４に記載の故障予測方法。     The second step includes
    Calculating a predicted value and a residual threshold of the amplitude value at a predetermined measurement time based on the slope;
    Storing the predicted value and the residual threshold in a memory;
    Calculating a residual that is a difference between the measured value of the amplitude value and the predicted value at the predetermined measurement time;
    The failure prediction method according to claim 23 or 24, wherein the measured value is excluded when the residual exceeds a range between a lower limit value and an upper limit value in the residual threshold.
PCT/JP2010/001865 2009-03-27 2010-03-16 Failure prediction system, electronic device, and failure prediction method WO2010109809A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2011505853A JP5573834B2 (en) 2009-03-27 2010-03-16 Failure prediction system, electronic device, and failure prediction method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2009-079079 2009-03-27
JP2009079079 2009-03-27

Publications (1)

Publication Number Publication Date
WO2010109809A1 true WO2010109809A1 (en) 2010-09-30

Family

ID=42780509

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2010/001865 WO2010109809A1 (en) 2009-03-27 2010-03-16 Failure prediction system, electronic device, and failure prediction method

Country Status (2)

Country Link
JP (1) JP5573834B2 (en)
WO (1) WO2010109809A1 (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011090104A1 (en) * 2010-01-21 2011-07-28 日本電気株式会社 Breakdown prediction device, breakdown prediction method, and breakdown prediction program
WO2013027742A1 (en) * 2011-08-23 2013-02-28 日本電気株式会社 Failure prediction method and failure prediction system for electronic apparatus
KR101307526B1 (en) * 2012-11-19 2013-09-12 (주)나다에스앤브이 Realtime vibration monitoring device
JP2015095094A (en) * 2013-11-12 2015-05-18 ルネサスエレクトロニクス株式会社 Semiconductor device, battery pack, and mobile terminal
CN109492833A (en) * 2018-12-25 2019-03-19 洛阳中科协同科技有限公司 A kind of bearing ring quality of production prediction technique based on decision Tree algorithms
JP2020092649A (en) * 2018-12-12 2020-06-18 株式会社クボタ Harvester
CN112394703A (en) * 2019-08-14 2021-02-23 中车时代电动汽车股份有限公司 Vehicle fault management system
CN114705285A (en) * 2022-04-01 2022-07-05 中国科学技术大学 Multi-channel vibration monitoring method
CN114755459A (en) * 2022-04-02 2022-07-15 中国北方车辆研究所 Method for monitoring state of rotating speed sensor based on residual error evaluation method
US11543799B2 (en) * 2019-01-16 2023-01-03 Buffalo Machinery Company Limited System and method for evaluating residual operating life of machine component

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0785410A (en) * 1993-09-14 1995-03-31 Toshiba Corp Apparatus and method for controlling head of magnetic disk apparatus
JPH11306748A (en) * 1998-04-16 1999-11-05 Nec Corp Disk array apparatus
JP2004063040A (en) * 2002-07-31 2004-02-26 Toshiba Corp Fault forecast system, disk device provided with fault forecasting function, fault forecasting method, and fault forecasting program

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0785410A (en) * 1993-09-14 1995-03-31 Toshiba Corp Apparatus and method for controlling head of magnetic disk apparatus
JPH11306748A (en) * 1998-04-16 1999-11-05 Nec Corp Disk array apparatus
JP2004063040A (en) * 2002-07-31 2004-02-26 Toshiba Corp Fault forecast system, disk device provided with fault forecasting function, fault forecasting method, and fault forecasting program

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011090104A1 (en) * 2010-01-21 2011-07-28 日本電気株式会社 Breakdown prediction device, breakdown prediction method, and breakdown prediction program
JPWO2011090104A1 (en) * 2010-01-21 2013-05-23 日本電気株式会社 Failure prediction apparatus, failure prediction method, and failure prediction program
US8891188B2 (en) 2010-01-21 2014-11-18 Nec Corporation Breakdown prediction device, breakdown prediction method, and breakdown prediction program
JP5712936B2 (en) * 2010-01-21 2015-05-07 日本電気株式会社 Failure prediction apparatus, failure prediction method, and failure prediction program
WO2013027742A1 (en) * 2011-08-23 2013-02-28 日本電気株式会社 Failure prediction method and failure prediction system for electronic apparatus
JPWO2013027742A1 (en) * 2011-08-23 2015-03-19 日本電気株式会社 Failure prediction method and failure prediction system for electronic equipment
KR101307526B1 (en) * 2012-11-19 2013-09-12 (주)나다에스앤브이 Realtime vibration monitoring device
US10096866B2 (en) 2013-11-12 2018-10-09 Renesas Electronics Corporation Semiconductor device, battery pack, and mobile terminal
JP2015095094A (en) * 2013-11-12 2015-05-18 ルネサスエレクトロニクス株式会社 Semiconductor device, battery pack, and mobile terminal
US10522882B2 (en) 2013-11-12 2019-12-31 Renesas Electronics Corporations Semiconductor device, battery pack, and mobile terminal with multi-speed CPU
JP2020092649A (en) * 2018-12-12 2020-06-18 株式会社クボタ Harvester
JP7090535B2 (en) 2018-12-12 2022-06-24 株式会社クボタ Harvester
CN109492833A (en) * 2018-12-25 2019-03-19 洛阳中科协同科技有限公司 A kind of bearing ring quality of production prediction technique based on decision Tree algorithms
US11543799B2 (en) * 2019-01-16 2023-01-03 Buffalo Machinery Company Limited System and method for evaluating residual operating life of machine component
CN112394703A (en) * 2019-08-14 2021-02-23 中车时代电动汽车股份有限公司 Vehicle fault management system
CN114705285A (en) * 2022-04-01 2022-07-05 中国科学技术大学 Multi-channel vibration monitoring method
CN114755459A (en) * 2022-04-02 2022-07-15 中国北方车辆研究所 Method for monitoring state of rotating speed sensor based on residual error evaluation method
CN114755459B (en) * 2022-04-02 2023-06-30 中国北方车辆研究所 Rotational speed sensor state monitoring method based on residual error assessment method

Also Published As

Publication number Publication date
JP5573834B2 (en) 2014-08-20
JPWO2010109809A1 (en) 2012-09-27

Similar Documents

Publication Publication Date Title
JP5573834B2 (en) Failure prediction system, electronic device, and failure prediction method
JP2010271073A (en) Diagnosis device of abnormality in equipment
CN110554092B (en) Detection system and detection method
JP5651998B2 (en) Abnormality diagnosis method and abnormality diagnosis system using pattern library
US6907368B2 (en) Method and device for detecting a pulse-type mechanical effect on a system part
US11635535B2 (en) Seismic sensor and earthquake determination method
US8891188B2 (en) Breakdown prediction device, breakdown prediction method, and breakdown prediction program
JP2007292489A (en) Insulation abnormality diagnosis system of electric facilities
JP2009281955A (en) Storage battery deterioration diagnostic system, storage battery deterioration diagnostic apparatus, storage battery deterioration diagnostic method and program
JP2013501927A (en) Early detection method of damage in vehicle transmission
WO2022166138A1 (en) Method and apparatus for measuring wear data of protective bearing
JP2009133810A (en) Vibration monitoring device
JP5203012B2 (en) Abnormality detection device, life detection device, abnormality detection method, and program for causing computer to function as abnormality detection device
JP4542918B2 (en) Bearing abnormality detection device
JP4943961B2 (en) Switchboard deterioration diagnosis system, switchboard deterioration diagnosis device, switchboard deterioration diagnosis method and program
JP2004063040A (en) Fault forecast system, disk device provided with fault forecasting function, fault forecasting method, and fault forecasting program
JP5203339B2 (en) Elevator rope monitoring device
JP5322742B2 (en) Reactor vibration monitoring method and reactor vibration monitoring system
JP4950861B2 (en) Information recording apparatus and control method of information recording apparatus
JP4585489B2 (en) Impact detection structure of information equipment
KR20140086269A (en) Method for predicting residual life of bearing
KR101035469B1 (en) Method and device for detecting a pulsed mechanical effect on a plant component
CN110657885B (en) Vibration alarming method and system of vibration sensor and terminal
JP6878057B2 (en) Insulation diagnostic equipment and insulation diagnostic method
JP2003083801A (en) Failure detection device and failure detection method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 10755613

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2011505853

Country of ref document: JP

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 10755613

Country of ref document: EP

Kind code of ref document: A1