CN101072984A - Vibration analysis system and method for a machine - Google Patents
Vibration analysis system and method for a machine Download PDFInfo
- Publication number
- CN101072984A CN101072984A CN 200580041688 CN200580041688A CN101072984A CN 101072984 A CN101072984 A CN 101072984A CN 200580041688 CN200580041688 CN 200580041688 CN 200580041688 A CN200580041688 A CN 200580041688A CN 101072984 A CN101072984 A CN 101072984A
- Authority
- CN
- China
- Prior art keywords
- data
- neural network
- vibration
- machinery
- input data
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Images
Abstract
A system and method for detecting and analyzing anomalies in a machine during operation. The system and method includes at least one sensor configured to detect characteristics of the machine indicative of machine vibration, at least one other sensor configured to detect characteristics of the machine indicative of other than machine vibration, a plurality of neural networks to receive input data, at least one neural network receiving vibration data from the at least one sensor, and at least one other neural network receiving non-vibration data from the at least one other sensor, and an expert system to receive output data from the neural networks and responsively analyze machine operation for anomalies.
Description
Technical field
Present invention relates in general to a kind of be used for analyzing machinery and the system and method vibration related data, relate in particular to a kind of system and method, use traditional vibration analysis technology in conjunction with artificial intelligence, test and analyze the vibration related data of the machinery of rotation.
Background technology
The machinery that rotates has a lot of application.For example, the machinery as removable machinery, as, road vehicle and offroad vehicle, building machinery, building machinery etc. utilize the principle of rotating to come work.Engine, motor, drive system, grounded parts such as wheel or track or the like make machinery carry out work by rotating.
Can analyze and determine the efficient and the life-span expectation of the machinery of rotation by the vibration in the research mechanical part.Friction force between the movable part is added the scrambling in the component tolerance, can cause the vibration in the machinery.Vibration analysis can help in real time and interference-free is determined the health condition of machinery, even can reach the stage of prediction component life and potential break.
Vibration analysis comprises the related notion of sound and ultrasonic analysis, mechanically makes us feeling interest for a long time aspect healthy monitoring and diagnose always.Yet it is defective that the vibration analysis technology is proved to be usually, they otherwise suspicious result is provided, otherwise provide the data that can not at once be explained or understand.
Attempted using artificial intelligence technology to test and analyze the vibration of machinery.For example, all transferred this assignee's U.S. Patent No. 5,566,092,5,566,273,5,602,761,5,854,993,6,236,950 and 6,539,319 have announced and use neural network to carry out the test and the analysis of machinery, particularly about the test and the analysis of the vibration characteristics of machinery, technical variation.Although these technology of implementing in above-mentioned patent have obtained success to a certain degree, yet still wish the further technology of exploitation, making it provides bigger reliability, robustness and precision in test with in analyzing.
The present invention is used for overcoming top illustrated one or more problems.
Summary of the invention
In one aspect of the invention, announced and be used to survey and analyze the unusual a kind of system of machinery at run duration.This system comprises that at least one is configured to survey the sensor of the mechanical vibration performance of described machinery, at least one is configured to survey other sensor of the on-mechanical vibration characteristics of described machinery, be used for receiving a plurality of neural networks of input data, at least one neural network receives vibration data from described at least one sensor, other neural network of at least one receives non-vibration data from described at least one other sensor, and one receives output data and analyzes the unusual expert system of mechanical movement responsively from described neural network.
In another aspect of the present invention, announced to be used to survey and analyze the unusual a kind of method of machinery at run duration.This method comprises receiving from a plurality of sensors imports data, at least one sensor is configured to survey the mechanical vibration performance of described machinery, at least one other sensor is configured to survey the on-mechanical vibration characteristics of described machinery, described input data are sent to a plurality of neural networks, at least one neural network receives vibration data, other neural network of at least one receives the non-vibration data, described at least one other neural network is used for contrasting the influence of described non-vibration data to described vibration data, send output data to an expert system from described a plurality of neural networks, and analyze described output data so that obtain relevant with mechanical movement unusual.
Of the present invention aspect another, announced a kind of system of the vibration that is used for surveying and analyze machinery.This system comprises a testing station, this station can receive the input data from a plurality of sensors, at least one sensor is configured to survey the mechanical vibration performance of described machinery, at least one other sensor is configured to survey the on-mechanical vibration characteristics of described machinery, this system also comprises a processing enter (hub), this center comprises a plurality of neural networks, at least one neural network can receive and analyze vibration data, other neural network of at least one receives and analyzes the non-vibration data, described processing enter also comprises an expert system, can also analyze the vibration characteristics of machinery from the data behind described a plurality of neural network receiving and analyzings responsively, and this system also comprises a communication linkage, can transmit data to described processing enter from described testing station.
Description of drawings
Fig. 1 is the synoptic diagram that can adopt a machinery of the present invention;
Fig. 2 is a block scheme, has described one aspect of the present invention; And
Fig. 3 is the detailed block scheme that is fit to the processing enter that the system in Fig. 2 uses.
Embodiment
With reference to accompanying drawing, shown a system 100 and a detection substantially and analyzed machinery 10 unusual methods at run duration.Unusually can refer to described mechanical 10 the characteristic relevant, refer in particular to the characteristic relevant of described mechanical 10 rotatable parts with vibration with vibration.
Particularly, with reference to figure 1, machinery 10 is embodied in a wheel loader, and it is generally used for earth work and building purposes.Described wheel loader describe only to be used for doing serve exemplary purposes, no matter various types of machineries that move or static, can adopt the present invention.For example, the wheel machine of other type and means of transport, track-mounted machinery, generator, be used for the machinery making, assemble and store and the machinery of various other types can benefit from use of the present invention.
Among Fig. 1 exemplary mechanical 10 comprises an engine 12, a variator (transmission) 14, one certain type transfer device (transfer unit) 16, for example, transmission box, and a power train (drive train) 18, all these parts all are well-known in the art.A common feature of described mechanical 10 the parts of listing above is, all comprises the rotation of element, and these parts tend to vibrate in acceptable margin tolerance or outside the scope.
Can detect vibration in the machinery 10 directly or by deriving with at least one vibration transducer 20.Vibration transducer 20 can comprise such device, and as rotational speed sensor, accelerometer etc., they can be connected each desirable position on described mechanical 10, detects the parameter that indication is rotated.
Can be other with at least one, non-vibration sensor 22 is surveyed directly or indirectly relevant with described mechanical 10 but relevant characteristic with vibration.The example of non-vibration sensor 22 has, temperature sensor, humidity sensor, barometer, liquid level sensor or the like.
Fig. 2 is a block scheme, has shown an aspect of described system 100.Described system 100 is depicted as has a telemetry station 102 and a processing enter 118 that perhaps is positioned at away from the central spot of described testing station 102 that is used for described mechanical 10 on-the-spot test.For example, many telemetry stations 102 can be arranged, these telemetry stations also are positioned at the position at mechanical place, and a single processing enter 118, and this processing enter is positioned at the central station away from mechanical position.Although content description of the present invention described testing station 102 and processing enter 118 be positioned at the situation at diverse location place, can imagine that described testing station 102 and processing enter 118 can be positioned at same position, even can be contained in the unit.
Described testing station 102 can comprise a plurality of sensing inputs 104, and these sensing inputs can be from for example various vibrations and non-vibration sensor 20,22.Described sensing input 104 can be transfused in the signal conditioner 106.Described signal conditioner 106 can be used to carry out such as suitable functions such as bias current are provided.Then, the described signal of having regulated can carry out anti-aliasing processing via analog filter 107.Then, described signal is sent to an I/O connector 108, and then delivers to an A/D converter card 109, so that be digital signal from analog signal conversion.Sensing is imported the signal that 104 a-m representative was handled by above-mentioned steps.Perhaps, importing the signal that 104 n-z are described as sensing can not need above-mentioned adjusting and filtering, and they can directly be delivered to I/O connector 108.
Processor 110 receives the data-signal of regulating its processing of Xingqi of going forward side by side.For example, storer 112 can be accessed, with storage and retrieve data.Described processor 110 for example by selecting some data, data being sorted out, the data tabulation is also partly analyzed raw data or the like, also can be ready to data so that give described processing enter 118.
Described telemetry station 102 is by power supply 114 power supplies, and this power supply can be a uninterrupted power supply (ups) Unity, avoids because the caused damage of power interruption with the data that protection equipment and institute receive and collect.
Communication linkage 116 provides communication between described testing station 102 and processing enter 118.Described communication linkage 116 can be wired, also can be wireless, and this depends on whole system configuration and needs.Preferably the way by the Internet protocol that is used for website visiting communicates.Yet, also can be with other communication means.For example, described communication linkage 116 can be wireless, such as radio, microwave or satellite communication, also can be wired, such as communicating by letter by telephone wire, coaxial cable, power transmission line or the like.In addition, described communication linkage 116 can make hard wired by special lead connection, perhaps, is joined together as an integrated unit between described testing station 102 and the processing enter 118.Use relevant transmitting and receiving apparatus (not shown) to be implemented in the communication of carrying out on the described communication linkage 116.No matter communicate, by can guarantee the safety in described communication linkage 116 visits such as the Internet protocol of encrypting by what medium.
Described processing enter 118 can receive data from one or more testing stations 102, handle and analyze these data, and the result that will analyze delivers to testing station 102, or deliver to the place of other appointment, exceed other place range of tolerable variance or alarm condition message as deliver to designated reception by note or Email.The details of operation of described processing enter 118 can give best description with reference to figure 3.Described processing enter 118 can be represented the software in the computing machine.In the present embodiment, the module among Fig. 3 is represented various software functions.
In Fig. 3, the test signal of processing enter 118 is delivered in the module representative that is labeled as time signal 302 by communication linkage 116.Described time signal 302 was handled by testing station 102 before transmitting, and can be numeral, encrypt from analog-converted, and for example transmit as the compression verification data with Internet protocol.Described time signal 302 can have many passages and for each passage many testing procedures be arranged.For example, described time signal 302 can have a plurality of passages, represent each analyzed transducing signal, and each passage can have a plurality of testing procedures, is used for described mechanical 10 multiple service condition.More specifically say, described mechanical 10 can be a variator, described passage can be represented from a plurality of sensors and receive the signal of coming, these sensors are configured to detect corresponding a plurality of characteristics of described variator, characteristic amplitude as vibration, temperature, pressure, environmental baseline or the like, and described testing procedure can be indicated the various operational modes of described variator, as first gear forward, second gear forward or the like.
A tunnel 304 is set at least comes time of reception signal 302 to send at least one algorithm 306 to.Described tunnel 304 can liken signal distributor (signal splitter) on function, therefore provide unique route for the data-signal 302 that will send.Yet described tunnel 304 does not arrive the physical connection in the external world of described processing enter 118.Time signal 302 can not add with changing passes through each tunnel 304, and perhaps, one or more tunnels 304 can be used to select the desirable component of described time signal 302.Described algorithm 306 can utilize such as association, covariance, wavelet analysis or the like and come processing time signal 302 to analyze.
Described hyperchannel selector switch 308 also can be for each the passage processing time signal 302 in the hyperchannel for example, so that be a neural network, such as No. three neural network 312, ready signal.Specifically, the value that described hyperchannel selector switch 308 can average signal is as peak value or power, so that neural network 312 provides a little for No. three.
Described time signal 302 also can be sent in an autoregression (AR) algorithm 314 so that be transformed in the frequency domain.AR algorithm 314 well known in the art is a kind of signal to be transformed into very effective method in the calculating of going the frequency domain from time domain.Then, the signal after the conversion can be sent in the neural network, in number one neural network 316, analyzes.
Each passage of described time signal 302 can be represented auxiliary sensor, that is, and and from the signal of non-vibration sensor 22.These aiding sensors signals can be sent in the compensation neural network, promptly in No. five neural network 318, are used for the influence of comparison auxiliary property to the analysis of vibration coherent signal.For example, such as temperature, pressure, liquid level, atmospheric conditions or the like auxiliary property the vibration of machinery had influence.Equally, if considered these auxiliary properties, the analysis of vibration signal can be carried out on a higher trust level.
Described time signal 302 also can be sent in FFT (fast fourier transform) control 320, and this control can be selected in the use of blocking FFT, extraction time (decimated time) waveform FFT etc.Such as number one FFT 322 and No. second FFT such as FFT 324 time-domain signal is converted to frequency-region signal, and can compresses or not compress the number of frequency domain point, so that simplify the work of neural network.
The use of fft algorithm can be handled time-domain signal so that produce frequency-region signal, and frequency-region signal has been provided by all frequency contents that exist in the time-domain signal that is provided.Consider lot of data point in the frequency spectrum, it is significant perhaps will spending definite which frequency of plenty of time, and determines that what is the normal level of these frequencies.The present invention can use a kind of self-template algorithm (autostencil algorithm), based on given frequency-region signal produce one group of bands of a spectrum (spectral divisions) automatically.Based on the yardstick of described bands of a spectrum, the new data signal can pass these bands of a spectrum and diagnose with regard to its content, such as the power of contained data in bare maximum that is included in any frequency content wherein or the bands of a spectrum, for example RMS value.
Preferred algorithm can come work like this,, discerns the critical frequencies composition in the described signal that is, produces bands of a spectrum around described signal, and the height control of described bands of a spectrum to suitable, is repeated this process up to producing a series of bands of a spectrum.Described algorithm can at first be discerned the frequency content of the high-amplitude of the part that do not belong to existing bands of a spectrum.Described algorithm utilizes a predefined or configurable transport function of user (transfer function) to produce bands of a spectrum then around this frequency content.Described transport function is used for discerning the width of described bands of a spectrum.Described transport function is an index curve normally.Because environmental variance, can cause moving of frequency as the variation of RPM, temperature, humidity etc., the frequency content that higher frequency content is lower is subjected to bigger influence usually.Described indicial transfer function can be used for widening those bands of a spectrum at upper frequency place, and its content is kept relative stability.A similar indicial transfer function can be used for adjusting the height of described bands of a spectrum, as, threshold value.Then, this transport function can be used to the threshold value of any bands of a spectrum in the lower frequency range is reduced in proportion, because it will take much better than power than at lower frequency place at higher frequency place makes the amplitude of frequency content produce significant the variation.
Can repeat to discern the process of frequency content, up to having satisfied certain standard.In first embodiment, need the bands of a spectrum of fixed number, promptly N is individual, discerns a N principal ingredient of described signal.In second embodiment, number N that can bands of a spectrum during the operation of described algorithm is set to infinity, so that cover described signal fully with bands of a spectrum, therefore, has guaranteed that all in fact signal contents are all covered by bands of a spectrum at least.The 3rd embodiment can comprise the general power of calculating signal to noise ratio (S/N ratio) and signal, begins an identification bands of a spectrum processing procedure then, and non-noise powers all in described signal all are comprised within the bands of a spectrum of definition.
FFT point from number one FFT can be directly used in a neural network, as in No. second neural network, to carry out vibration analysis.
FFT point from No. second FFT 324 can be sent to a series of bandpass filter (BPF) 326 subsequently, and these wave filter configured in parallel make frequency domain can be divided into a plurality of frequency band bags.Each bag can be further processed by frequency packet handler 328 subsequently.For example, can determine r.m.s. (RMS) value of each bag, perhaps can determine the peak value (PK) of each bag.Perhaps, can determine power spectrum density (PSD) or some other this values.
Point from described frequency packet handler 328 can be used to a neural network subsequently, as in No. four neural network 332, analyzes tested mechanical 10 vibration characteristics.In addition, these points also can be used for No. five neural network 318 in addition simultaneously or only be used for neural network 318 alternatively No. five, and promptly described compensation neural network is so that comprise the influence of auxiliary property in tested mechanical 10 vibration analysis.Preferably, owing to comprised the factor that the common vibration analysis technology that can exert an influence to the vibration of machinery does not have consideration, the result of described compensation neural network will have higher trust level.
During the initial setting up of described processing enter 118, and during the described processing enter 118 of fine tuning subsequently is with the acquisition optimum performance, can be each used described neural network and adjustable weight factor of algorithm assigns.Be provided with in the process of described processing enter 118 and can use guide (wizard), this some input of guide prompting input is such as number and kind, BPF details, the tunnel that will use and algorithm or the like of neural network.Therefore, final processing enter configuration can be customized, to be used for concrete vibration analysis situation, also periodically change as required.Described processing enter 118 can be based on software, but this software provides user's allocative abilities of drag and drop.For example, various selectable characteristics, such as BPF, algorithm, neural network or the like, can by drag represent these characteristics icon to certain position of display, put down this icon then and select these characteristics.By this technology, can select any requisite number purpose characteristic.Can carry out described characteristic for each testing procedure of each passage in the described time signal 302 selects.
Above-mentioned algorithm, and other also spendable algorithm can be used to discern a series of inputs that may be used in the neural network.The described layoutprocedure of part comprises, accepts collected data, it is confirmed as a normal mode, and train this normal mode, to be suitable in all neural networks that described processing enter 118 comprised.When a new test plan was set, perhaps when upgrading or revising an existing test plan, each neural network of current existence must can reconfigure based on collected latest data potentially.As the part of this process, each neural network of current existence must be reinitialized, produced, trained.Subsequently, must confirm the normal mode of each neural network, and train each neural network with a new pattern.
Therefore can make this process full automation with described guide, just eliminate operator's error, and reduce significantly and make described processing enter 118 enter the required time of operation.Described guide can be discerned each neural network of current existence, and no matter the previous state of any neural network how.In addition, the use of guide allows to carry out the test identification of normal specimens, and sample promptly is whole results of a test being undertaken by described processing enter 118.This can comprise the identification of a plurality of samples, and the pattern that these samples have subsequently can be combined to form single input pattern.In case confirmed an input pattern for each unique neural network, this network just is cleared (if necessary), is initialised, is trained then, to have the normal mode of being confirmed by said process.Repeat this process for each existing neural network then, and no matter the size of described pattern, for example, the number of point.
Industrial applicibility
Described system 100 can be selected in any one pattern in several patterns and move. For example, Selectable first pattern is, described system 100 monitors all test plan parameters that are configured As a result, if the threshold value of any selected parameter is exceeded, then output warning. Can be with notice with any uncommon The mode of hoping is delivered to any desirable place.
Second pattern can comprise the characteristic of above-mentioned first pattern, and comprises that further any presetting can Reliability is exceeded the output warning function based on neutral net in the situation.
The 3rd pattern can comprise the characteristic of described first and second patterns, and further comprises a kind of energy Power so that the Test Engineer can select to make described system 100 based on any selected parameter weighting result with And any neutral net confidence level is carried out intelligent diagnostics. Make up based on the weight in any desired category These results can be provided for described mechanical 10 operator, the test worker of processing center 118 Cheng Shi, the database that records, other appointed place or the combination of any above-mentioned situation.
In described processing enter 118, use above-mentioned technology and parts can make the result have higher confidence level, and make it possible to utilize neural network to carry out vibration analysis, and these neural networks do not need historical data in the past to obtain the reliability and the determinacy of reasonable level yet.
Can from the research of accompanying drawing, disclosure and appended claims book, obtain other aspect.
Claims (9)
1. be used for detection and analyze machinery (10) unusual a kind of system (100), comprising at run duration:
At least one sensor (20) is configured to survey the characteristic of the mechanical vibration of described machinery (10);
At least one other sensor (22) is configured to survey the characteristic of the on-mechanical vibration of described machinery (10);
A plurality of neural networks (312,316,318,330,332), be used for receiving the input data, at least one neural network (312,316,318,330,332) receives vibration data from described at least one sensor (20), at least one other neural network (318) receives non-vibration data from described at least one other sensor (22), and described at least one other neural network (318) is used for the influence of more described non-vibration data to described vibration data; And
An expert system (334) receives output data and analyzes the unusual of mechanical movement responsively from described neural network (312,316,318,330,332).
2. system according to claim 1 (100), wherein, at least one neural network comprises weight factor in described a plurality of neural networks (312,316,318,330,332).
3. system according to claim 1 (100) also comprises an autoregression algorithm (314), is used for receiving the input data of time domain, and in frequency domain, and the data after will changing offer at least one neural network (316) with described input data-switching.
4. according to system according to claim 1 (100), also comprise at least one fast fourier transform (322,324), be used for receiving the input data, described input data are transformed into frequency domain from time domain, and the data after will changing offer at least one neural network (318,330,332).
5. system according to claim 1 (100), also comprise a hyperchannel selector switch (308), be used for from described input data, selecting needed each data channel, the passage that treatment of selected is selected, give a related algorithm (310) analyzing the interaction between described each passage selected channel transfer, and selected passage is offered at least one neural network (312).
6. be used for detection and analyze machinery (10) unusual a kind of method, comprise step at run duration:
Receive the input data from a plurality of sensors (20,22), at least one sensor (20) is configured to survey the characteristic of the mechanical vibration of described machinery (10), and at least one other sensor (22) is configured to survey the characteristic of the on-mechanical vibration of described machinery (10);
Transmit described input data to a plurality of neural networks (312,316,318,330,332), at least one neural network (312,316,318,330,332) receives vibration data, at least one other neural network (318) receives non-vibration data, and described at least one other neural network (318) is used for the influence of more described non-vibration data to described vibration data;
The output data of described a plurality of neural networks (312,316,318,330,332) is sent to an expert system (334); And
Analyze described output data so that obtain relevant with mechanical movement unusual.
7. method according to claim 6, wherein, the output data that transmits described a plurality of neural network (312,316,318,330,332) is included as from the step of the data allocations weight factor of at least one neural network (312,316,318,330,332) output.
8. method according to claim 6, wherein, transmit described input data and comprise step for a plurality of neural networks (312,316,318,330,332):
The described input data of at least a portion are transformed into frequency domain from time domain; And
Transmit described frequency domain data at least one neural network (316,318,330,332).
9. method according to claim 6, wherein, transmit described input data and comprise step for a plurality of neural networks (312,316,318,330,332):
From described input data, select desirable data channel;
Utilize the interaction between selected each passage of related algorithm (310) analysis; And
The data channel of being analyzed is sent at least one neural network (312).
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US63348104P | 2004-12-06 | 2004-12-06 | |
| US60/633,481 | 2004-12-06 | ||
| US11/197,707 | 2005-08-04 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101072984A true CN101072984A (en) | 2007-11-14 |
Family
ID=38899482
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200580041688 Pending CN101072984A (en) | 2004-12-06 | 2005-10-17 | Vibration analysis system and method for a machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101072984A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108680244A (en) * | 2018-04-26 | 2018-10-19 | 浙江大学 | A kind of rotating machinery vibrating wireless monitoring device and method |
| CN109346198A (en) * | 2018-09-18 | 2019-02-15 | 深圳中广核工程设计有限公司 | A kind of fuel for nuclear power plant clad failure diagnostic system and its diagnostic method |
| CN111106034A (en) * | 2018-10-28 | 2020-05-05 | 台湾积体电路制造股份有限公司 | Annealing apparatus and method |
| CN112204362A (en) * | 2018-05-30 | 2021-01-08 | 西门子工业软件公司 | Method and apparatus for detecting vibration and/or acoustic transmission in mechanical systems |
| CN112513760A (en) * | 2018-08-14 | 2021-03-16 | 西门子股份公司 | Device and method for predicting the remaining service life of a machine |
-
2005
- 2005-10-17 CN CN 200580041688 patent/CN101072984A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108680244A (en) * | 2018-04-26 | 2018-10-19 | 浙江大学 | A kind of rotating machinery vibrating wireless monitoring device and method |
| CN108680244B (en) * | 2018-04-26 | 2020-06-09 | 浙江大学 | Rotary machine vibration wireless monitoring device and method |
| CN112204362A (en) * | 2018-05-30 | 2021-01-08 | 西门子工业软件公司 | Method and apparatus for detecting vibration and/or acoustic transmission in mechanical systems |
| CN112204362B (en) * | 2018-05-30 | 2022-10-04 | 西门子工业软件公司 | Method and apparatus for detecting vibration and/or acoustic transmission in mechanical systems |
| CN112513760A (en) * | 2018-08-14 | 2021-03-16 | 西门子股份公司 | Device and method for predicting the remaining service life of a machine |
| CN109346198A (en) * | 2018-09-18 | 2019-02-15 | 深圳中广核工程设计有限公司 | A kind of fuel for nuclear power plant clad failure diagnostic system and its diagnostic method |
| CN111106034A (en) * | 2018-10-28 | 2020-05-05 | 台湾积体电路制造股份有限公司 | Annealing apparatus and method |
| US11587807B2 (en) | 2018-10-28 | 2023-02-21 | Taiwan Semiconductor Manufacturing Co., Ltd. | Annealing apparatus and method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7194383B2 (en) | Vibration analysis system and method for a machine | |
| US7225108B2 (en) | Cross correlation diagnostics tool for vibration analysis | |
| CN102192825B (en) | User interface system and method for diagnosing a rotating machine condition not based upon prior measurement history | |
| EP0148256B1 (en) | System for operational monitoring of a machine | |
| US7187773B2 (en) | System for evaluating abnormal sound, sound recorder and apparatus for evaluating abnormal sound | |
| US6053047A (en) | Determining faults in multiple bearings using one vibration sensor | |
| US5995910A (en) | Method and system for synthesizing vibration data | |
| US5493288A (en) | System for multidirectional information transmission between at least two units of a drilling assembly | |
| US6289735B1 (en) | Machine diagnostic system and method for vibration analysis | |
| CN101072984A (en) | Vibration analysis system and method for a machine | |
| EP1192421B1 (en) | A non-invasive system and method for diagnosing potential malfunctions of semiconductor equipment components | |
| CN102308194B (en) | An analysis system | |
| CN100549636C (en) | The method and apparatus of detection shafting vibration on motor | |
| WO1997015044A1 (en) | Acoustic signature analysis for a noisy environment | |
| US20030194946A1 (en) | Grinding wheel system | |
| IL33261A (en) | Process and installation for measuring the physical characteristics of rocks during drilling | |
| CN104081173A (en) | Apparatus and method for analysing the condition of a machine having a rotating part | |
| CN105424364A (en) | Diagnostic method and device of train bearing failure | |
| CN101636589A (en) | Method for the detection of errors in pump units | |
| US7444265B2 (en) | Machine and/or monitoring | |
| CN104048825A (en) | Multi-sensor fusion type fault locating method for gearbox of wind turbine generator | |
| CN102713539A (en) | Method and device for monitoring torsional vibrations of a rotary shaft of a turbine engine | |
| WO1997010580A1 (en) | Apparatus for wireless transmission of shaft position information | |
| CN205300880U (en) | Transformer winding vibration manipulation test system | |
| Kozochkin et al. | Vibration testing of technological processes in automated machining equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| AD01 | Patent right deemed abandoned | ||
| C20 | Patent right or utility model deemed to be abandoned or is abandoned |