CN112557036A - Diagnostic apparatus and method, and computer-readable storage medium - Google Patents

Abstract

The present invention relates to a diagnostic device and method, and a computer-readable storage medium for diagnosing a deterioration condition of a bearing mechanism. The diagnostic device (10) comprises: an acquisition unit (32) that acquires measurement data relating to vibration corresponding to rotation of a bearing mechanism that includes a rolling element between an outer wheel and an inner wheel; an extraction unit (34) that extracts a feature amount from the result of frequency analysis of the measurement data; an estimation unit (36) that estimates the size of the scratch occurring on the outer wheel or the inner wheel based on a predetermined relationship between the change in the characteristic amount and the size of the scratch occurring on the outer wheel or the inner wheel, and the characteristic amount extracted by the extraction unit (34); and an output unit (38) that outputs the estimation result obtained by the estimation unit (36).

Description

Diagnostic apparatus and method, and computer-readable storage medium

Technical Field

The present invention relates to a diagnostic apparatus, a diagnostic method, a diagnostic program, and a computer-readable storage medium.

Background

Conventionally, for preventive maintenance of a production machine using a reduction gear or the like including a bearing mechanism, periodic inspection for diagnosing a deterioration state of the reduction gear or the like and determining whether or not replacement is required has been performed. When deterioration such as scratches occurs in the bearing mechanism, iron powder may be mixed into the lubricating oil or grease (grease) in the bearing mechanism. Therefore, in the periodic maintenance, an operator usually stops the production machine, opens the outer cover, and takes out the lubricating oil or grease in the bearing mechanism to measure the iron powder concentration.

As a technique related to such diagnosis, for example, a robot control device has been proposed which can perform abnormality diagnosis of a robot without preparation in advance and without interrupting a production process. The robot control device acquires first data used in the abnormality diagnosis in time series, and acquires second data used to extract the first data used in the abnormality diagnosis in time series. The robot control device extracts first data corresponding to the extraction time of the first data used for the abnormality diagnosis, which is specified based on the second time-series data, and executes the abnormality diagnosis of the robot based on the extracted first data (see patent document 1).

[ Prior art documents ]

[ patent document ]

[ patent document 1] Japanese patent laid-open No. 2016-179527

Disclosure of Invention

[ problems to be solved by the invention ]

The apparatus described in patent document 1 diagnoses an abnormality of a motor or a reduction gear based on data that can be easily collected such as a motor current without stopping a target production machine (robot). However, the device described in patent document 1 can only diagnose whether or not there is an abnormality, and cannot grasp the deterioration condition of the bearing mechanism. Therefore, there is a problem that the user cannot determine when what kind of countermeasure should be taken.

The present invention has been made in view of the above, and an object thereof is to provide a diagnostic device, a method, and a program that can diagnose a deterioration condition of a bearing mechanism.

[ means for solving problems ]

To achieve the above object, a diagnostic device according to the present invention includes: an acquisition unit that acquires data relating to vibration corresponding to rotation of a bearing mechanism that includes a rolling element between an outer ring and an inner ring; an extraction unit that extracts a feature amount from a result of frequency analysis of the data acquired by the acquisition unit; an estimating unit that estimates a size of a scratch occurring in the outer wheel or the inner wheel based on a predetermined relationship between a change in the characteristic amount and a size of a scratch occurring in the outer wheel or the inner wheel and the characteristic amount extracted by the extracting unit; and an output unit that outputs the estimation result obtained by the estimation unit.

The extraction unit may extract, as the feature value of the outer wheel, an amplitude of a frequency predetermined as a frequency at which a scratch occurs in the outer wheel, and may extract, as the feature value of the inner wheel, an amplitude of a frequency predetermined as a frequency at which a scratch occurs in the inner wheel, among results of frequency analysis of the data.

The estimating unit may estimate that a scratch is generated on the outer wheel when the characteristic value of the outer wheel exceeds a predetermined threshold value, and estimate that a scratch is generated on the inner wheel when the characteristic value of the inner wheel exceeds the threshold value.

In addition, the estimating unit may estimate that, when it is estimated that the outer ring has a scratch, a scratch size at a time point when the characteristic amount that changes with time reaches the nth maximum value is a size that is (n-1/2) times the half cycle of the outer ring circumference, a scratch size at a time point when the characteristic amount that changes with time reaches the nth minimum value is a size that is n times the half cycle of the outer ring circumference, or a scratch size at a time point when the characteristic amount that changes with time reaches the nth maximum value is a size that is (n-1/2) times the distance between the rolling elements, and a scratch size at a time point when the characteristic amount that changes with time reaches the nth minimum value is a size that is n times the distance between the rolling elements.

The estimating unit may estimate, when it is estimated that the inner ring has a scratch, a scratch size at a time point when the characteristic amount that changes with time reaches a first maximum value and a scratch size at a time point when a difference between an nth maximum value and an n-1 st maximum value of the characteristic amount that changes with time is smaller than a predetermined value is a size (n-1/2) times a distance between the rolling elements, estimate a scratch size at a time point when the characteristic amount that changes with time reaches an nth minimum value as a size n times the distance between the rolling elements, and estimate that two scratches have occurred in the inner ring when the difference between the nth maximum value and the n-1 st maximum value of the characteristic amount that changes with time is equal to or larger than the predetermined value.

The estimating unit may estimate that, when two scratches are estimated to be generated in the inner wheel, the size of one of the scratches is larger than the size of the scratch estimated last time, and the size of the other scratch is smaller than the size of the one scratch.

The estimating unit may estimate the sizes of the two scratches at the time point when the characteristic amount that changes with time reaches the nth maximum value, based on a combination in which a difference between the size of the one scratch and the size of the other scratch is n times the distance between the rolling elements, and estimate the sizes of the two scratches at the time point when the sum of the size of the one scratch and the size of the other scratch is n times or 1/2 times the distance between the rolling elements.

The estimating unit may estimate a maximum value or an average value of the combination as the sizes of the two scratches.

The estimating unit may estimate the size of the scratch at the time when the nth maximum value or minimum value is reached based on the temporal change in the feature amount, and estimate the sizes of the two scratches based on the combination of values in a predetermined range including the estimated sizes.

The extraction unit may normalize the extracted feature amount using the first feature amount and the second feature amount corresponding to the rotation speed of the bearing mechanism when the data is acquired by the acquisition unit, or a predetermined value and the second feature amount, which are predetermined as values corresponding to the first feature amount, which are determined based on a predetermined relationship between the rotation speed of the bearing mechanism and the first feature amount in a state where no scratches are present and the second feature amount in a state where predetermined scratches are present.

The extraction unit may normalize the extracted feature amount by setting the first feature amount or the predetermined value as a minimum value and the second feature amount as a maximum value. Further, the extraction unit may be configured to use, as the second feature amount, a value within a predetermined range including a first maximum value of the feature amount that changes with time.

Further, the diagnostic method of the present invention includes: an acquisition unit acquires data relating to vibration corresponding to rotation of a bearing mechanism including a rolling element between an outer ring and an inner ring; an extraction unit that extracts a feature amount from a result of frequency analysis of the data acquired by the acquisition unit; an estimation unit that estimates a size of a scratch generated in the outer wheel or the inner wheel based on a predetermined relationship between a change in the characteristic amount and the size; and estimating the size of the scratch generated on the outer wheel or the inner wheel based on the feature value extracted by the extraction unit, and outputting the estimation result obtained by the estimation unit by an output unit.

Furthermore, a computer-readable storage medium of the present invention stores a diagnostic program for causing a computer (computer) to function as: an acquisition unit that acquires data relating to vibration corresponding to rotation of a bearing mechanism that includes a rolling element between an outer ring and an inner ring; an extraction unit that extracts a feature amount from a result of frequency analysis of the data acquired by the acquisition unit; an estimating unit that estimates a size of a scratch occurring in the outer wheel or the inner wheel based on a predetermined relationship between a change in the characteristic amount and a size of a scratch occurring in the outer wheel or the inner wheel and the characteristic amount extracted by the extracting unit; and an output unit that outputs the estimation result obtained by the estimation unit.

[ Effect of the invention ]

According to the diagnosis device, method and computer-readable storage medium of the present invention, the deterioration condition of the bearing mechanism can be diagnosed.

Drawings

Fig. 1 is a block diagram showing a hardware configuration of a diagnostic apparatus.

Fig. 2 is a diagram for explaining a relationship between the diagnostic apparatus and the device to be diagnosed.

Fig. 3 is a view showing a schematic configuration of a bearing (bearing).

Fig. 4 is a diagram showing a schematic configuration of a wave gear device.

Fig. 5 is a diagram showing an example of measurement data.

Fig. 6 is a block diagram showing an example of a functional configuration of the diagnostic apparatus.

Fig. 7 is a diagram for explaining the relationship between the diagnostic apparatus and the device to be diagnosed.

Fig. 8 is a diagram for explaining the relationship between the outer wheel scratch frequency and the rotation frequency.

Fig. 9 (a) to 9 (D) are diagrams for explaining the relationship between the measurement data and the scratches generated on the inner wheel or the outer wheel.

Fig. 10 (a) to 10 (D) are diagrams for explaining measurement data when the rolling element continuously passes through the scratch generated in the inner ring or the outer ring.

Fig. 11 (a) to 11 (D) are diagrams for explaining the relationship between the scratch generated in the outer wheel and the measurement data.

Fig. 12 is a diagram for explaining two scratches generated in the wave gear device.

Fig. 13 (a) to 13 (C) are diagrams for explaining the relationship between the measurement data and the scratch generated in the inner wheel.

Fig. 14 is a diagram showing an example of the feature DB.

Fig. 15 is a graph showing changes in the outer wheel scratch characteristic amount with respect to the operating time.

Fig. 16 is a graph showing changes in the inner wheel scratch characteristic amount with respect to the operation time.

Fig. 17 is a diagram showing an example of the diagnosis result DB.

Fig. 18 is a flowchart showing an example of the diagnosis process in the first embodiment.

Fig. 19 is a flowchart showing an example of the estimation processing.

Fig. 20 is a diagram showing an example of a model showing a relationship between the rotation speed and the characteristic amount of the bearing mechanism.

Fig. 21 is a flowchart showing an example of the diagnosis process in the second embodiment.

[ description of symbols ]

10. 210: diagnostic device

12：CPU

14: memory device

16: storage device

18: input device

20: output device

22: storage medium reading device

24: communication I/F

26: bus line

32: acquisition unit

34. 234: extraction section

36: estimation unit

38: output unit

42: characteristic value DB

44: diagnosis result DB

60: diagnostic object apparatus

62: motor with a stator having a stator core

64: speed reducer

66: vibration sensor

72: outer wheel

74: inner wheel

76: rolling body

S10, S20, S30, S35, S40, S60, S70, S41 to S50: step (ii) of

Detailed Description

Hereinafter, an example of an embodiment of the present invention will be described with reference to the drawings. In the drawings, the same or equivalent constituent elements and portions are denoted by the same reference numerals. For convenience of explanation, the sizes and ratios shown in the drawings are exaggerated and may be different from the actual ratios.

< first embodiment >

Fig. 1 is a block diagram showing a hardware configuration of a diagnostic device 10 according to a first embodiment. As shown in fig. 1, the diagnostic apparatus 10 includes a Central Processing Unit (CPU) 12, a memory 14, a storage device 16, an input device 18, an output device 20, a storage medium reading device 22, and a communication Interface (Interface) 24. The components are communicatively connected to each other via a bus 26.

The storage device 16 stores a diagnostic program for executing diagnostic processing. The CPU 12 is a central processing unit and executes various programs or controls the respective components. That is, the CPU 12 reads out the program from the storage device 16 and executes the program with the memory 14 as a work area. The CPU 12 performs control of the respective configurations and various arithmetic processes in accordance with programs stored in the storage device 16.

The Memory 14 includes a Random Access Memory (RAM) as an operation area for temporarily storing programs and data. The storage device 16 includes a Read Only Memory (ROM) and a Hard Disk Drive (HDD) or a Solid State Drive (SSD), and stores various programs including an operating system and various data.

The input device 18 is a device for performing various inputs, such as a keyboard and a mouse. The output device 20 is a device for outputting various information, such as a display or a printer. The output device 20 may also function as the input device 18 by using a touch panel display.

The storage medium reading device 22 reads data stored in various storage media such as a Compact Disc-Read Only Memory (CD-ROM), a Digital Versatile Disc-Read Only Memory (DVD-ROM), a blu-ray Disc (blue-ray Disc), and a Universal Serial Bus (USB) Memory, and writes data to the storage media.

The communication I/F24 is an Interface for communicating with other devices, and for example, standards such as Ethernet (registered trademark), Fiber Distributed Data Interface (FDDI), and Wi-Fi (registered trademark) are used.

As shown in fig. 2, the diagnostic apparatus 10 diagnoses scratches by using a production machine such as a robot as a diagnostic device 60. The apparatus 60 to be diagnosed includes a motor 62 and a speed reducer 64.

The motor 62 operates (rotates) a shaft of a bearing mechanism provided in the reduction gear 64 in accordance with an operation diagram (profile) generated by a robot control device (not shown) based on an operation command input by a user. The operation map refers to an acceleration/deceleration map, a motion map, and the like, and refers to characteristics and conditions such as the speed and acceleration/deceleration of the operation of the motor 62. For example, the operation diagram is shown as a change in speed with respect to time, such as trapezoidal acceleration and deceleration.

The speed reducer 64 includes a bearing mechanism. In the present embodiment, a wave gear device widely used as a speed reducer of a robot will be described as a main example of a bearing mechanism.

As shown in fig. 3, a general bearing mechanism (bearing) includes rolling elements 76 between an outer ring 72 and an inner ring 74, and the ball bearing (shape of the inner ring 74) is a perfect circle. On the other hand, as shown in fig. 4, the wave gear device is similar to a general bearing in that rolling elements 76 are provided between an outer ring 72 and an inner ring 74, but the ball bearing is elliptical. Thus, the periodicity of the vibration is different from that of a general bearing mechanism. In the deterioration diagnosis of the wave gear device, there are many points common to a general bearing mechanism, but in the present embodiment, diagnosis is performed focusing on the difference in the periodicity of the vibration.

Here, a basic principle of deterioration diagnosis by vibration, which targets a bearing mechanism, will be described.

Fig. 5 shows an example of measurement data relating to vibration corresponding to rotation of the bearing mechanism, such as output torque of the motor 62 or shaft vibration. When the outer ring 72 or the inner ring 74 of the bearing mechanism is scratched, the rolling elements 76 vibrate each time they pass through the scratched portion, and as shown in fig. 5, they show periodic changes in the measured data according to the shaft rotation speed. Therefore, the presence or absence of the scratch can be diagnosed by performing frequency analysis such as Fast Fourier Transform (FFT) on the measurement data and analyzing whether or not a peak of the spectrum or the size of the peak is present at a specific frequency corresponding to each of the outer wheel 72 and the inner wheel 74.

Next, a functional configuration of the diagnostic device 10 according to the first embodiment will be described.

Fig. 6 is a block diagram showing an example of the functional configuration of the diagnostic apparatus 10. As shown in fig. 6, the diagnostic apparatus 10 includes an acquisition unit 32, an extraction unit 34, an estimation unit 36, and an output unit 38 as functional components. In a predetermined storage area of the diagnostic apparatus 10, a feature Database (Database, DB)42 and a diagnostic result DB 44 are stored. Each functional configuration is realized by the CPU 12 reading out the diagnostic program stored in the storage device 16, expanding the diagnostic program in the memory 14, and executing the program.

The acquisition unit 32 acquires measurement data, which is data relating to vibration according to rotation of the bearing mechanism and is used for monitoring the state of the bearing mechanism, from the diagnosis target apparatus 60. The measurement data can be easily collected data such as motor current. In the example of fig. 2, the case where the output torque indicated by the motor current value is acquired as the measurement data of the motor 62 is shown, but the present invention is not limited to this. For example, as shown in fig. 7, a vibration sensor 66 for detecting vibration of the speed reducer 64 may be provided, and the acquisition unit 32 may acquire sensor output from the vibration sensor 66 as measurement data, or may acquire other data such as encoder data as measurement data. The acquisition unit 32 delivers the acquired measurement data to the extraction unit 34.

The extraction unit 34 performs frequency analysis on the measurement data delivered from the acquisition unit 32 by FFT or the like, and extracts a feature amount from the analysis result. Specifically, the extracting unit 34 extracts, as a feature amount corresponding to the scratch generated in the outer wheel 72 (hereinafter referred to as "outer wheel scratch feature amount"), an amplitude of a frequency predetermined as a frequency at which a peak occurs when the scratch is generated in the outer wheel 72, from a result of frequency analysis of the measurement data. The extracting unit 34 extracts an amplitude of a frequency predetermined as a frequency at which a scratch occurs in the inner wheel 74 as a feature amount corresponding to the scratch occurring in the inner wheel 74 (hereinafter referred to as "inner wheel scratch feature amount").

The outer wheel scratch characteristic amount and the inner wheel scratch characteristic amount are described in more detail.

In a bearing in which the inner ring 74 is formed in a perfect circle, the inner ring 74 rotates, but the outer ring 72 is fixed. The structure of the speed reducer 64 and the rotation frequency f of the motor 62_rotThe frequency f of vibration caused by the rolling elements 76 passing the inner race 74 at the scratches (hereinafter referred to as "inner race scratch frequency") thereof_inThe compound is determined by the following formula (1). The frequency f of vibration caused by the rolling elements 76 passing over the outer ring 72 scratches (hereinafter referred to as "outer ring scratch frequency") is_outDetermined by the following formula (2).

[ number 1]

Here, R₁Radius of the inner wheel 74, R₂The radius of the outer race 72 (see fig. 3), N the number of rolling elements 76, and C the reduction ratio (0 in the case of a bearing).

In the wave gear device, the inner wheel 74 has an elliptical shape. In the wave gear device, the inner ring 74 rotates at the rotational speed of the motor 62, and the outer ring 72 rotates in reverse at the speed of C (reduction ratio). Therefore, in the wave gear device, the same theory as that of the bearing is used for the inner wheel scratch frequency f_inAnd outer wheel scratch frequency f_outThis is true and the scratch generated at the outer wheel 72 is even at twice the rotational frequency (2 f)_rot) Can also be observed.

Specifically, in the wave gear device, as shown in fig. 4, since a strong force is always applied in the longitudinal direction, the inner ring 74 is likely to be scratched in the longitudinal direction. As shown in the upper diagram of fig. 8, the rolling elements 76 vibrate each time they pass through a scratch appearing in the longitudinal direction, and the frequency of the vibration substantially matches the above equation (1).

On the other hand, as shown in the middle part of fig. 8, the outer wheel 72 is rotated only at the rotational frequency f_rotThe inner ring 74 rotates once and receives a strong force when it is abutted in the longitudinal direction. Therefore, as shown in the lower diagram of fig. 8, the rolling elements 76 vibrate only when they pass through the notches and the long axis direction of the inner ring 74 passes through the notches. Thus, at the outer wheel scratch frequency f_outIt is somewhat difficult to observe the vibration, on the other hand, even at twice the rotation frequency (2 f)_rot) The vibration caused by the scratch can be observed.

Based on the above, the extracting unit 34 extracts the inner wheel scratch frequency f shown in the following table from the result of the frequency analysis_inAnd outer wheel scratch frequency f_outThe respective amplitudes (power values of the spectrum) are used as the inner wheel scratch characteristic amount and the outer wheel scratch characteristic amount.

[ Table 1]

In addition, as the outer ring scuffing frequency in the wave gear device, two f shown in this table may be used_outAny one of these methods may be used. Further, it is also possible to monitor two frequencies and use either one of the frequencies based on the size of the extracted outer wheel scratch feature amount. In addition, compare 2f_rotIn the case of (2), f_outThe state of the scratch can be finely monitored. In the following embodiments, the term "f" is used_out＝2f_rotThe case of the outer wheel scratch frequency will be described.

The extracting unit 34 delivers the outer wheel scratch feature value and the inner wheel scratch feature value extracted from the result of frequency analysis of the measurement data for a fixed amount of time to the estimating unit 36 at predetermined time intervals (for example, 10 minutes to 1 hour).

The estimating unit 36 estimates which of the outer wheel 72 and the inner wheel 74 has the scratch based on a predetermined relationship between the change in the characteristic amount and the size of the scratch generated in the outer wheel 72 or the inner wheel 74 and the characteristic amount passed from the extracting unit 34. When any one of the outer wheel 72 and the inner wheel 74 is scratched, the estimation unit 36 estimates the size of the scratch generated in the outer wheel 72 or the inner wheel 74.

Here, in order to explain the relationship between the characteristic amount and the size of the scratch, first, the vibration when the rolling element 76 passes through the scratch portion, that is, the change of the measurement data, is explained. The size of the measurement data has a relationship with the depth of the scratch, but the following description will be given assuming that the depth of the scratch is constant.

As shown in fig. 9 (a), when no scratch is generated on the outer wheel 72 or the inner wheel 74, the measurement data is flat data. As shown in fig. 9 (B), when a dot-like mark is formed on the outer ring 72 or the inner ring 74, the value of the measurement data greatly changes as long as the rolling element 76 passes through the marked portion. As shown in fig. 9 (C) and 9 (D), when a planar scratch having a size X [ mm ] is generated in the outer ring 72 or the inner ring 74, the value of the measurement data greatly changes at the time when the rolling element 76 rolls from a portion having no scratch into the scratched portion. Hereinafter, this change is also referred to as "roll-in vibration". At the time point when the rolling element 76 rolls out from the scratched portion to the non-scratched portion, the measurement data greatly changes on the side opposite to the roll-in vibration. This variation is also referred to as "roll-out vibration" hereinafter.

Fig. 10 (a) to 10 (D) show the case where the rolling elements 76 continuously pass through the scratched portion in the case of the same scratch as in fig. 9 (a) to 9 (D). The distance between the rolling bodies is set to L [ mm ].

In the case of fig. 10 (a), the measurement data is flat data as in the case of fig. 9 (a). In the case of fig. 10 (B), the value of the measurement data continuously changes as long as the rolling elements 76 pass through the respective scratched portions. As shown in fig. 10 (C), when the size X [ mm ] of the scratch is much smaller than the distance L [ mm ], specifically, when X < L/2, the rolling-in vibration of one rolling element 76 and the rolling-out vibration of the other rolling element 76 are continuously generated without being cancelled out by each other. When X is close to L/2, rolling-in vibration and rolling-out vibration occur at equal intervals due to the respective rolling elements 76 coming in and out of the scratch. Therefore, in the case of performing frequency analysis, the amplitude at a specific frequency is maximized.

As shown in fig. 10 (D), when X approaches L, rolling-in vibration of one rolling element 76 and rolling-out vibration of the other rolling element 76 tend to cancel each other out. In particular, when X is equal to L, the roll-in vibration of one rolling element 76 and the roll-out vibration of the other rolling element 76 are timed to coincide with each other, and therefore the measurement data are canceled out, and when frequency analysis is performed, the amplitude of a specific frequency is minimized.

Next, a change in measurement data when the outer wheel 72 is scratched in the wave gear device will be described. As shown in fig. 11 (a) to 11 (D), the major axis of the elliptical inner ring 74 is defined as a major axis a on one side and a major axis B on the other side, and the outer ring circumference is defined as L at 1/2_OUT. In fig. 11 a to 11D, the measurement data is shown by being divided into data showing the vibration when the major axis a passes through the scratch position (hereinafter referred to as "measurement data of the major axis a") and data showing the vibration when the major axis B passes through the scratch position (hereinafter referred to as "measurement data of the major axis B"). The actual measured measurement data is obtained by adding the two.

As shown in fig. 11 (a), when the long axis a passes through the position of the scratch and the rolling element 76 passes through the scratch portion at the start of the scratch, the measurement data of the long axis a is largely changed and the measurement data of the long axis B is flat. As shown in FIG. 11 (B), the scratch is extended, and X ≦ L for the size X of the scratch_OUTThe range of/2 is that the timing of the roll-in scratch and the roll-out scratch is deviated between the major axis a and the major axis B, and therefore the roll-in vibration and the roll-out vibration do not coincide with each other in the measurement data of the major axis a and the measurement data of the major axis B. As shown in FIG. 11 (C), the scratch is further extended at L_OUT/2＜X≦L_OUTSince the timings of the roll-in scratch and the roll-out scratch coincide with each other for the major axis a and the major axis B, the roll-in vibration and the roll-out vibration coincide with each other, and the tendency of the measurement data of the major axis a to cancel out the measurement data of the major axis B increases. As shown in FIG. 11 (D), the scratch is further extended at L_OUT＜X≦3L_OUTThe range of/2, roll-in scratch and roll-out for major axis A and major axis BThe timing of the scratch is reversed, and the measurement data has a waveform similar to that in fig. 11 (B).

Next, a change in measurement data when the inner ring 74 is scratched in the bearing mechanism will be described. The case where the scratch is one place is the same as the basic case shown in fig. 9 (a) to 9 (D) and fig. 10 (a) to 10 (D).

As a premise, it is assumed that the scratch is first generated one and becomes two at a time. As shown in fig. 12, the two scratches are located at substantially opposite positions in the longitudinal direction. The scratch generated first is referred to as a scratch a, and the scratch generated later is referred to as a scratch B. In the case where the scratch appearance positions are completely opposed to each other and the number of the rolling elements 76 is an odd number, the phase of the change in the measurement data corresponding to the rolling of the rolling elements 76 into each of the scratch a and the scratch B is shifted by 180 degrees.

In fig. 13 a to 13C, the measurement data is shown by being divided into data indicating the vibration when the rolling element 76 passes the scratch a (hereinafter, referred to as "measurement data of the scratch a") and data indicating the vibration when the rolling element 76 passes the scratch B (hereinafter, referred to as "measurement data of the scratch B"). The actual measured measurement data is obtained by adding the two.

As shown in fig. 13 (a), when one scratch is formed, the measurement data of the scratch a largely changes and the measurement data of the scratch B becomes flat when the rolling element 76 passes through the portion of the scratch a. As shown in fig. 13 (B), when the timing of the rolling of one rolling element 76 into the scratch a matches the timing of the rolling of the other rolling element 76 into the scratch B, the rolling vibrations overlap, and therefore the measurement data is largely changed by adding the two. The same applies to the case where the timing at which one rolling element 76 rolls out of the scratch a coincides with the timing at which the other rolling element 76 rolls out of the scratch B. In the case where the former and the latter occur simultaneously, the change in the measurement data reaches the maximum. At this time, in the result of the frequency analysis, the amplitude at a specific frequency is maximized.

On the other hand, as shown in fig. 13 (C), when the timing of the rolling-in of one rolling element 76 into the scratch a and the timing of the rolling-out of the other rolling element 76 from the scratch B coincide with each other, the rolling-in vibration and the rolling-out vibration overlap each other, and therefore the measurement data are canceled out. The same applies to the case where the timing at which one rolling element 76 rolls out of the scratch a coincides with the timing at which the other rolling element 76 rolls into the scratch B. In the case where the former and the latter occur simultaneously, the change in the measurement data is minimized. At this time, in the result of the frequency analysis, the amplitude at a specific frequency is minimized.

Similarly, when the timing of the rolling-in of a certain rolling element 76 into the scratch a coincides with the timing of the rolling-out of another rolling element 76 from the scratch a, the rolling-in vibration and the rolling-out vibration overlap, and therefore the measurement data cancel out. The same applies to the case where the rolling-out of one rolling element 76 from the scratch B coincides with the timing of the rolling-in of the other rolling element 76 into the scratch B. In the case where the former and the latter occur simultaneously, the change in the measurement data is minimized. At this time, in the result of the frequency analysis, the amplitude at a specific frequency is minimized.

The above-mentioned contents are further specifically explained. The centers of the two scratches are located at opposite positions in the longitudinal direction of the inner ring 74, and the waveforms of the measurement data of the scratch a and the scratch B when the scratches are dotted are expressed by the following expressions (3) and (4).

Scratch A ═ ag (f)_in,0)…(3)

Scratch B ═ bg (f)_in,π)…(4)

g (f,0) is a periodic function of the frequency f of the zero-phase, pulsed waveform, and g (f, θ) is a periodic function of the frequency f with a phase of θ [ rad ]. And, a and b are the amplitude of the periodic waveform.

Generation of X_A[mm]The periodic waveform of the scratch a at the time of the scratch is the sum of the roll-in vibration and the roll-out vibration, and is expressed by the following formula (5).

Scratch A ═ ag (f)_in,-θ_A)-ag(f_in,+θ_A)…(5)

Here, it is assumed that the roll-in vibration and the roll-out vibration have opposite phases. That is, if the distance between the rolling elements is set to L_IN[mm]When X is present_A＝L_INIs θ_AAs shown in the following equation (6), the roll-in vibration and the roll-out vibration cancel each other and become zero.

Scratch A ═ ag (f)_in,-π)-ag(f_in,+π)…(6)

Otherwise, it is expressed as θ_A＝πX_A/L_INTherefore, the periodic waveform of the scratch a is expressed by the following formula (7).

Scratch A ═ ag (f)_in,-πX_A/L_IN)-ag(f_in,+πX_A/L_IN)…(7)

Likewise, generating X_B[mm]The periodic waveform of the scratch B at the time of scratching is expressed by the following formula (8).

Scratch B ═ bg (f)_in,π-πX_B/L_IN)-bg(f_in,π+πX_B/L_IN)＝bg(f_in,π(1-X_B/L_IN))-bg(f_in,π(1+X_B/L_IN))…(8)

According to the expressions (7) and (8), two scratches are generated at the opposite positions in the longitudinal direction of the inner wheel 74, and when the depths of the two scratches are the same, the maximum value or the minimum value of the inner wheel scratch characteristic is obtained under the conditions shown in the following table.

[ Table 2]

Based on this, the estimating unit 36 estimates the presence or absence of the scratch and the size of the scratch as follows.

The estimation unit 36 stores the feature amount delivered from the extraction unit 34 in the feature amount DB 42 shown in fig. 14 in association with the operation time of the diagnosis target device 60. Then, the estimation unit 36 reads the outer wheel scratch feature amount per operation time stored in the feature amount DB 42, and draws a change in the outer wheel scratch feature amount with respect to the operation time of the diagnosis target device 60 as shown in fig. 15. Similarly, the estimation unit 36 reads the inner wheel scratch feature amount per operation time stored in the feature amount DB 42, and draws a change in the inner wheel scratch feature amount with respect to the operation time of the diagnosis target apparatus 60 as shown in fig. 16.

The estimation unit 36 estimates that the outer ring 72 has been scratched when the outer ring scratch feature value exceeds a predetermined threshold value for the first time in a change in the outer ring scratch feature value shown in fig. 15, and performs estimation processing assuming that there is a scratch in the outer ring 72 at the time of a subsequent diagnosis. Similarly, the estimation unit 36 estimates that the inner ring 74 has a scratch when the inner ring scratch feature value exceeds a predetermined threshold value for the first time in the change of the inner ring scratch feature value shown in fig. 16, and performs the estimation process on the assumption that the inner ring 74 has a scratch at the time of the subsequent diagnosis.

That is, the characteristic value is the inner wheel scratch frequency f_inWhether the amplitude (inner wheel scratch characteristic quantity) of (D) is changed or the outer wheel scratch frequency f_outThe amplitude (outer wheel scratch characteristic amount) of the vibration is changed, and it can be discriminated whether the scratch is generated at the inner wheel or the outer wheel.

In order to estimate whether or not a scratch has occurred, the average and dispersion of the outer wheel scratch characteristic amount and the inner wheel scratch characteristic amount at the normal time may be calculated in advance, and the average + dispersion value may be used as the threshold value. Further, a position where the increase rate from the average value in each normal state is large may be estimated as the position of the scratch for the inner wheel scratch characteristic amount and the outer wheel scratch characteristic amount.

When it is estimated that the outer wheel 72 is scratched, the estimating unit 36 estimates the size of the scratch at the time point when the outer wheel scratch characteristic amount shown in fig. 15 reaches the nth maximum value to be a size (n-1/2) times the half-cycle amount of the outer wheel circumference. The estimating unit 36 estimates the size of the scratch at the time point when the outer wheel scratch characteristic reaches the nth minimum value to be n times the size of the half circumference of the outer wheel circumference.

If L is provided_OUTWhen the outer ring circumference is 0.5 × in the example of fig. 15, the size of the outer ring scratch is estimated as follows at each time point of (a) to (D).

(A) Time point of maximum value of the first time (n ═ 1): the size of the scratch of the outer wheel is 0.5L_OUT

(B) Time point of the first minimum value (n ═ 1): the size of the scratch of the outer wheel is 1.0L_OUT

(C) Time point of maximum value of the second time (n ═ 2): the size of the scratch of the outer wheel is 1.5L_OUT

(D) Time point of minimum value of the second time (n ═ 2): the size of the scratch of the outer wheel is 2.0L_OUT

When it is estimated that the inner ring 74 has a scratch, the estimating unit 36 estimates the size of the scratch at the time point when the inner ring scratch characteristic amount shown in fig. 16 reaches the first maximum value to be a size (n-1/2) times the distance between the rolling elements. The estimating unit 36 estimates the size of the scratch at the time point when the difference between the nth maximum value and the n-1 st maximum value of the inner ring scratch feature amount is smaller than the predetermined value to be (n-1/2) times the distance between the rolling elements in the same manner. Further, the estimating unit 36 estimates the size of the scratch at the time point when the inner ring scratch characteristic amount reaches the nth minimum value to be a size n times the distance between the rolling elements.

The estimating unit 36 estimates that two scratches are generated in the inner wheel 74 when the difference between the nth maximum value and the n-1 st maximum value of the inner wheel scratch feature amount is equal to or greater than a predetermined value. At this time, the estimating unit 36 estimates that the size of the scratch a (the scratch generated first) is equal to or larger than the size of the scratch at the time of the previous estimation, and estimates that the size of the scratch B (the scratch generated later) is smaller than the size of the scratch a.

When the scratch is one place, the nth maximum value and the (n-1) th maximum value do not change greatly, but when the scratch is two places, the nth maximum value becomes larger than the (n-1) th maximum value. The predetermined value for estimating whether the scratch is one or two may be a predetermined value of the above-mentioned divisible value or a predetermined ratio (for example, 50%) of the n-1 th maximum value.

Specifically, the estimating unit 36 estimates the sizes of the two scratches at the time point when the inner ring scratch feature value reaches the nth maximum value, based on the combination of the candidate sizes in which the difference between the size of the scratch a and the size of the scratch B is n times the distance between the rolling elements. The estimating unit 36 estimates the sizes of the two scratches at the time point when the inner ring scratch feature value reaches the nth minimum value, based on a combination of candidate sizes in which the sum of the size of the scratch a and the size of the scratch B is n times or n/2 times the distance between the rolling elements.

More specifically, presumeThe part 36 will scratch the size X of the A_ADimension X of scratch B_BThe maximum value of the sum is the value of the inner wheel circumference, X according to the conditions of the above Table 2_AAnd X_BThe combination of (2) is set as a combination of the candidate sizes.

Furthermore, the estimating unit 36 may estimate X_AAnd X_BThe maximum value of the sum is the size of the scratch at the time point when the nth maximum value or minimum value is reached, which is predicted based on the change in the past inner wheel scratch characteristic amount. For example, assume Δ T from operating time T1 to operating time T2_1-2In the period (2), the size of the scratch becomes large by Δ X_1-2At the time point of the operation time T2, the size of the scratch is X_T2. In this case, the scratch size X at the time of the operating time T3 can be predicted as shown in the following equation (9)_T3。

X_T3＝X_T2+(ΔX_1-2×ΔT_2-3)/ΔT_1-2

In this case, the estimating unit 36 can estimate X_AAnd X_BThe maximum value of the sum is set to X_T3X will be according to the conditions of said Table 2_AAnd X_BThe combination of (2) is set as a combination of the candidate sizes. This can suppress the combination of the candidate sizes from becoming enormous. In addition, X may be_AAnd X_BThe maximum value of the sum is set as X_T3The margin value is added.

The estimating unit 36 estimates, for example, the maximum value or the average value of the combination of the candidate sizes as the sizes of the scratch a and the scratch B.

If the distance L between the rolling bodies is set_IN＝8[mm]In the example of fig. 16, the size of the inner wheel scratch is estimated at each of the time points (E) to (I) as follows.

(E) Time point of maximum value of the first time (n ═ 1):

the size of the inner wheel scratch of … is 0.5L_IN＝4mm

(F) Time point of the first minimum value (n ═ 1):

the size of the inner wheel scratch of … is 1.0L_IN＝8mm

(G) The maximum value of the second time (n is 2) and the difference between the maximum values of (E) is less than a predetermined value:

the size of the inner wheel scratch of … is 1.5L_IN＝12mm

(H) Time point of minimum value of the second time (n ═ 2):

the size of the inner wheel scratch of … is 2.0L_IN＝16mm

(I) A third maximum value (n is 3) and a difference between the maximum values of (G) is equal to or greater than a predetermined value:

the scratch is estimated from the following combination of candidate sizes at two points …

In the case of (I), the time periods (E) to (H) were about 600 hours, and the size of the scratch A was increased by 12mm during the time periods (H) to (I) were about 600 hours. Using this, the size X of the scratch A at the time point of (I)_ADimension X of scratch B_BThe maximum value of the sum is set to X at the time point of 12 mm. times.2 (two scratches) +16mm ((H))_A) 40 mm. And, X_AThe size of the time point of (H) is 16mm or more, X_BLess than X_AAnd according to condition "X" of Table 2_A-X_B＝nL_IN", is X_A-X_B3 × 8-24 mm. Therefore, when the candidate size is 1mm, the following combinations (X)_A，X_B) Each becomes a candidate size.

(25，1)、(26，2)、(27，3)、(28，4)、(29，5)、(30，6)、(31，7)、(32，8)

If the maximum value is selected from the combination of the candidate sizes, X is estimated_AIs 32mm, X_BIs 8 mm. Then, if the average value is taken, X is estimated_AIs 28.5mm, X_BIs 4.5 mm.

In addition, in the case of f using the above formula (2)_outWhen the outer wheel scratch frequency is set, the estimating unit 36 can estimate the size of the outer wheel scratch in the same manner as the size of the inner wheel scratch.

The estimation unit 36 stores the estimated position of the scratch (whether the outer wheel 72 or the inner wheel 74) and the size of the scratch as a diagnosis result in the diagnosis result DB 44 shown in fig. 17, for example, in association with the operation time.

The output section 38 outputs the diagnosis result stored in the diagnosis result DB 44. The output diagnosis result may be only the latest diagnosis result or may be output together with the past diagnosis result. The output method can be any method such as displaying a screen on a display, outputting a sound from a speaker, and printing a ticket by a printer.

Next, the operation of the diagnostic device 10 according to the first embodiment will be described.

Fig. 18 is a flowchart showing a flow of the diagnosis process executed by the CPU 12 of the diagnosis apparatus 10. The CPU 12 reads out the diagnostic program from the storage device 16, expands the diagnostic program in the memory 14, and executes the program, whereby the CPU 12 functions as each functional component of the diagnostic device 10 and executes the diagnostic process shown in fig. 18.

In step S10, the acquisition unit 32 determines whether or not a predetermined time (for example, 10 minutes to 1 hour) has elapsed since the previous execution of the diagnosis process. If the determination has been made, the process proceeds to step S20, and if not, the determination of this step is repeated.

In step S20, the acquisition unit 32 acquires, from the diagnosis target apparatus 60, measurement data such as the output torque of the motor 62 as data relating to vibration according to the rotation of the bearing mechanism.

Next, in step S30, the extraction unit 34 performs frequency analysis on the measurement data acquired by the acquisition unit 32 by FFT or the like. Then, the extracting unit 34 extracts the inner wheel scratch frequency f shown in table 1 from the result of the frequency analysis_inAnd outer wheel scratch frequency f_outThe respective amplitudes are defined as inner wheel scratch characteristic values and outer wheel scratch characteristic values. The extracting unit 34 delivers the extracted outer wheel scratch feature and inner wheel scratch feature to the estimating unit 36. The estimation unit 36 stores the feature amount delivered from the extraction unit 34 in the feature amount DB 42 shown in fig. 14 in association with the operation time of the diagnosis target device 60.

Next, in step S40, the estimation processing is executed. Here, the estimation processing will be described in detail with reference to fig. 19.

In step S41, the estimation unit 36 refers to the diagnosis result DB 44 to determine whether or not the diagnosis result of the scratch is stored. If both the inner wheel scratch feature value and the outer wheel scratch feature value are maintained at normal values, the diagnosis result DB 44 does not store the diagnosis result of the scratch. If the diagnosis result of the scratch is stored, the process proceeds to step S44, and if not, the process proceeds to step S42.

In step S42, the estimation unit 36 determines whether or not any of the inner wheel scratch feature value and the outer wheel scratch feature value exceeds a threshold value. If the threshold value is exceeded, the process proceeds to step S43, and if not, the estimation process is ended, and the process returns to the diagnosis process (fig. 18).

In step S43, the estimation unit 36 stores the position of the scratch in the inner wheel 74 when the inner wheel scratch feature value exceeds the threshold value, and the position of the scratch in the outer wheel 72 when the outer wheel scratch feature value exceeds the threshold value, in the diagnosis result DB 44 in association with the operation time of the diagnosis target device 60. Then, the estimation processing is ended, and the processing returns to the diagnosis processing (fig. 18).

In step S44, the estimation unit 36 determines whether or not a new maximum value or minimum value has occurred in any of the change in the inner wheel scratch feature amount and the change in the outer wheel scratch feature amount. If a new maximum value or minimum value is generated, the process proceeds to step S45, and if not, the estimation process is terminated, and the process returns to the diagnosis process (fig. 18).

In step S45, the estimation unit 36 determines whether the inner wheel scratch feature value or the outer wheel scratch feature value that has generated the new maximum value or minimum value. If the outside wheel scratch feature value is the outside wheel scratch feature value, the process proceeds to step S46, and if the inside wheel scratch feature value is the inside wheel scratch feature value, the process proceeds to step S47.

In step S46, if it is determined in step S44 that the generated new maximum value or minimum value is the nth maximum value of the outer wheel scratch feature amount, the estimation unit 36 sets (n-1/2) × L_OUT(L_OUT0.5 x outer wheel circumference) is estimated as the size of the outer wheel scratch. If it is determined in step S44 that the generated new maximum value or minimum value is the nth minimum value of the outer wheel scratch feature value, the estimation unit 36 sets n × L_OUTPresumably the size of the outer wheel scratch.

In step S47, the estimation unit 36 determines whether the inner wheel scratch feature amount is the first maximum value or whether the difference between the nth maximum value and the (n-1) th maximum value of the inner wheel scratch feature amount is smaller than a predetermined value, thereby determining whether the inner wheel scratch is one spot. If the inner wheel scratch is one, the process proceeds to step S48, and if there are two, the process proceeds to step S49. In addition, referring to the diagnostic result DB 44, if two inner wheel scratches are estimated in the past diagnostic result, the determination in this step may be omitted and the routine may proceed to step S49.

In step S48, if it is determined in step S44 that the new maximum value or minimum value generated is the nth maximum value of the inner wheel scratch feature amount, the estimation unit 36 sets (n-1/2) × L_IN(L_INDistance between rolling elements) is estimated as the size of the inner ring scratch. If it is determined in step S44 that the new maximum value or minimum value is the nth minimum value of the inner wheel scratch feature value, the estimation unit 36 sets n × L_INPresumably the size of the inner wheel scratch.

In step S49, the estimating unit 36 estimates the size X of the scratch a (scratch generated earlier)_AThe size X of the scratch B (the scratch generated later) is set to be larger than the size of the scratch in the previous estimation_BIs set to be less than X_A. Then, the estimating unit 36 outputs X_AAnd X_BThe maximum value of the sum is a value predicted based on the inner wheel circumference or the past inner wheel scratch feature amount, and X is determined according to the conditions in table 2_AAnd X_BThe combination of (2) is set as a combination of the candidate sizes. The estimating unit 36 estimates the maximum value or the average value of the combination of the candidate sizes as X_AAnd X_B。

Next, in step S50, the estimation unit 36 stores the position of the scratch estimated in step S45 (the outer wheel 72 or the inner wheel 74) and the size of the scratch estimated in step S46, step S48, or step S49 in the diagnostic result DB 44 in association with the operation time as diagnostic results. Then, the estimation unit 36 delivers the diagnosis result to the output unit 38. Then, the estimation processing is ended, and the processing returns to the diagnosis processing (fig. 18).

Next, in step S60 of the diagnosis process (fig. 18), the output unit 38 outputs the diagnosis result stored in the diagnosis result DB 44.

Next, in step S70, the output unit 38 determines whether or not the diagnosis result has reached the end condition of the diagnosis process. As the termination condition, for example, a condition that the size of the scratch has reached a size set in advance by the user, a condition that two scratches are generated, or the like is predetermined. If the end condition is not met, the process returns to step S10, and if the end condition is met, the diagnostic process is ended.

As described above, according to the diagnostic device of the first embodiment, the inner wheel scratch frequency f is extracted from the result of the frequency analysis of the measurement data on the vibration according to the rotation of the bearing mechanism_inAnd outer wheel scratch frequency f_outThe respective amplitudes are used as the inner wheel scratch characteristic amount and the outer wheel scratch characteristic amount. The size of the inner wheel scratch or the outer wheel scratch is estimated based on a predetermined relationship between the change in the inner wheel scratch characteristic amount and the outer wheel scratch characteristic amount and the size of the scratch. This makes it possible to determine not only whether the bearing mechanism is abnormal, but also the position and size of the scratch occurring in the bearing mechanism, thereby diagnosing the deterioration of the bearing mechanism.

Further, since the deterioration state of the bearing mechanism can be diagnosed as described above, the user can determine the timing of the periodic inspection, the reduction gear replacement, the replacement of the lubricating oil, and the like in accordance with the deterioration state, and easily determine the countermeasure and the countermeasure timing in accordance with the deterioration state of the device to be diagnosed. Further, since information indicating the detailed deterioration state, i.e., the size of the scratch, is obtained, it is also easy to take a measure for temporarily extending the lifetime of the device to be diagnosed, for example, a measure such as a low-load operation, in accordance with the delivery date of the replacement part.

Further, since the operation of measuring the concentration of the iron powder by taking out the lubricant oil or the like in the reduction gear is not required, the man-hour for the diagnosis operation itself can be reduced.

< second embodiment >

Next, a second embodiment will be explained. In the diagnostic apparatus according to the second embodiment, the same components as those of the diagnostic apparatus 10 according to the first embodiment are denoted by the same reference numerals, and detailed description thereof is omitted. The hardware configuration of the diagnostic apparatus according to the second embodiment is the same as that of the diagnostic apparatus 10 according to the first embodiment shown in fig. 1, and therefore, the description thereof is omitted.

As shown in fig. 2 or 7, the diagnostic apparatus 210 according to the second embodiment diagnoses scratches by using a production machine such as a robot as the diagnostic device 60, as in the diagnostic apparatus 10 according to the first embodiment.

As shown in fig. 6, the diagnostic device 210 includes the acquisition unit 32, the extraction unit 234, the estimation unit 36, and the output unit 38 as functional components. In addition, the feature DB 42 and the diagnosis result DB 44 are stored in a predetermined storage area of the diagnosis device 210. Each functional configuration is realized by the CPU 12 reading out the diagnostic program stored in the storage device 16, expanding the diagnostic program in the memory 14, and executing the program.

The extracting unit 234 extracts the outer wheel scratch feature amount and the inner wheel scratch feature amount, as in the extracting unit 34 of the first embodiment.

Here, when the apparatus 60 to be diagnosed is a robot that performs conveyor sorting (conveyor picking) or the like, the operation speed and the operation distance of the robot are different from cycle to cycle, and thus the rotation speed of the bearing mechanism is also different from cycle to cycle. In this case, the accuracy of estimating the scratch size may be lower even if the state of the scratch in the bearing mechanism is the same as compared with the case of the diagnosis target device 60 that performs a simple operation. This is because the manner of fluctuation of the acquired measurement data differs depending on the operation of the apparatus 60 to be diagnosed from that in the case of the apparatus 60 to be diagnosed that is simply operated.

Therefore, even if the operation of the diagnosis target device 60 differs from cycle to cycle, the extraction unit 234 according to the second embodiment can stably extract the outer wheel scratch feature value and the inner wheel scratch feature value while suppressing the influence of the estimation of the scratch size due to the fluctuation of the measurement data.

Specifically, the extracting unit 234 normalizes the extracted outer wheel scratch feature value and inner wheel scratch feature value to suppress variation in measurement data due to the operation of the diagnosis target device 60. More specifically, the extracting unit 234 normalizes the outer wheel scratch feature value and the inner wheel scratch feature value so that the value of the feature value extracted in the normal state without scratches becomes the minimum value and the first maximum value in the change of the feature value (fig. 15 and 16) described in the first embodiment becomes the maximum value. The reason why the first local maximum value is used is that the first local maximum value at which the feature amount changes is likely to be scratched, and the size of the feature amount is likely to be fixed. However, if the size of the scratch is known to some extent, normalization may be performed so that an arbitrary nth maximum value becomes a maximum value. The feature amount extracted in the normal state without scratches is not limited to the case of using the minimum value in the normalization, and a predetermined value (for example, zero) may be used as a value corresponding to the feature amount extracted in the normal state without scratches. The maximum value itself is not limited to the case of using the maximum value itself as the maximum value in the normalization, and a value within a predetermined range including the maximum value, that is, a value in the vicinity of the maximum value may be used. For example, a value smaller than the maximum value by a predetermined value may be normalized as a maximum value.

For example, the maximum value and the minimum value are obtained in advance for each rotational speed of the bearing mechanism according to the operation assumed for the device 60 to be diagnosed, and as shown in fig. 20, a model in which the rotational speed is associated with the maximum value and the minimum value of the feature amount is prepared in advance. The model can be created based on the characteristic amount extracted from the measurement data by measuring the measurement data while varying the rotation speed of the bearing mechanism in each of the normal state where no scratch is present and the state where a scratch having a predetermined size is generated in an experimental environment. In addition, although fig. 20 shows a model for normalizing the inner wheel scratch feature value, the same model may be prepared for the outer wheel scratch feature value.

Here, the thicker the lubricating oil film in the speed reducer 64, the smaller the vibration of the speed reducer 64. The thickness of the lubricating oil film in the speed reducer 64 is proportional to the rolling speed of the rolling elements 76. Therefore, as shown in fig. 20, in the model showing the relationship between the rotation speed and the characteristic amount of the bearing mechanism, when the rotation speed becomes equal to or higher than a predetermined speed, the thickness of the lubricating oil film in the reduction gear 64 increases in accordance with the rotation speed, and the characteristic amount, which is the vibration caused by the scratches, becomes small (a in fig. 20). On the other hand, when the rotation speed of the bearing mechanism is extremely low, a lubricating oil film is hardly formed. That is, since the thickness of the lubricating oil film is close to zero, the vibration, i.e., the characteristic amount, becomes larger as the rotation speed of the bearing mechanism increases (B in fig. 20). As described above, the model can be said to be expressed by two models of an increase in vibration associated with an increase in the rotational speed of the bearing mechanism and a decrease in vibration associated with an increase in the thickness of the lubricating oil film caused by an increase in the rotational speed of the bearing mechanism.

The extraction unit 234 acquires the rotation speed of the bearing mechanism at the time of extracting the feature amount, and acquires the maximum value and the minimum value of the feature amount at the rotation speed with reference to the model. Then, the extraction unit 234 normalizes the feature values by, for example, the following expression (3) using the maximum value and the minimum value of the acquired feature values.

[ number 2]

In formula (3), x'_tIs a normalized feature quantity, x_tIs a feature quantity before normalization, v, extracted from measurement data measured at time t_tIs the rotational speed of the bearing mechanism at time t. And, f_max(v_t) And f_min(v_t) Respectively obtained from the model, and a rotational speed v_tThe maximum and minimum values of the associated feature quantity.

Next, the operation of the diagnostic device 210 according to the second embodiment will be described.

Fig. 21 is a flowchart showing a flow of the diagnosis process executed by the CPU 12 of the diagnosis apparatus 210. The CPU 12 reads out the diagnostic program from the storage device 16, expands the diagnostic program in the memory 14, and executes the program, whereby the CPU 12 functions as each functional component of the diagnostic device 210 and executes the diagnostic process shown in fig. 21.

The diagnostic process in the second embodiment is different from the diagnostic process in the first embodiment (fig. 18) in that step S35 is performed between step S30 and step S40. In step S35, the extracting unit 234 normalizes the outer wheel scratch feature value and the inner wheel scratch feature value extracted in step S30, respectively, using the model shown in fig. 20 and the expression (3) described above, for example, and delivers the normalized feature values to the estimating unit 36.

As described above, according to the diagnostic device of the second embodiment, the maximum value and the minimum value of the feature amount corresponding to the rotation speed of the bearing mechanism at the time of feature amount extraction are acquired based on the model which is prepared in advance and shows the relationship between the rotation speed of the bearing mechanism and the feature amount. Then, the extracted feature amount is normalized using the acquired maximum value and minimum value. Thus, the influence of the fluctuation of the measurement data due to the operation of the diagnostic target apparatus is suppressed, and the same effect as that of the first embodiment can be obtained even when the diagnostic target apparatus having different operation speeds and operation distances for each cycle is targeted.

In the above embodiments, the bearing mechanism is mainly described as an example of a wave gear device, but a general bearing is also an application range of the present invention. In this case, when it is determined that the inner wheel scratch is present in step S45 of the estimation process (fig. 19), the process proceeds to step S48, and the processes of step S47 and step S49 may be omitted.

In the above embodiments, the diagnostic process executed by the CPU reading the software (program) may be executed by various processors other than the CPU. Examples of the processor in this case include a dedicated electric Circuit, which is a processor having a Circuit configuration designed specifically to execute a Specific process, such as a Programmable Logic Device (PLD) such as a Field-Programmable Gate Array (FPGA) whose Circuit configuration can be changed after manufacture, an Application Specific Integrated Circuit (ASIC), and the like. The diagnostic process may be executed by one of these various processors, or may be executed by a combination of two or more processors of the same type or different types (e.g., a plurality of FPGAs, a combination of a CPU and an FPGA, or the like). More specifically, the hardware configuration of these various processors is an electric circuit in which circuit elements such as semiconductor elements are combined.

In the above embodiments, the description has been given of the form in which the diagnostic program is stored (installed) in the storage device in advance, but the present invention is not limited to this. The program may be provided in a form stored in a storage medium such as a CD-ROM, a DVD-ROM, a blu-ray disc, or a USB memory. The program may be downloaded from an external device via a network.

Claims (15)

1. A diagnostic device comprising:

an acquisition unit that acquires data relating to vibration corresponding to rotation of a bearing mechanism that includes a rolling element between an outer ring and an inner ring;

an extraction unit that extracts a feature amount from a result of frequency analysis of the data acquired by the acquisition unit;

an estimating unit that estimates a size of a scratch occurring in the outer wheel or the inner wheel based on a predetermined relationship between a change in the characteristic amount and the size of the scratch occurring in the outer wheel or the inner wheel and the characteristic amount extracted by the extracting unit; and

and an output unit that outputs the estimation result obtained by the estimation unit.

2. The diagnostic device of claim 1, wherein

The extraction unit extracts, as a feature value of the outer wheel, an amplitude of a frequency predetermined as a frequency at which a scratch occurs in the outer wheel, and extracts, as a feature value of the inner wheel, an amplitude of a frequency predetermined as a frequency at which a scratch occurs in the inner wheel, among results of frequency analysis of the data.

3. The diagnostic device of claim 2, wherein

The estimation unit estimates that a scratch is generated in the outer wheel when the characteristic value of the outer wheel exceeds a predetermined threshold value, and estimates that a scratch is generated in the inner wheel when the characteristic value of the inner wheel exceeds the threshold value.

4. The diagnostic device of claim 3, wherein

When the estimation unit estimates that the outer wheel has a scratch,

the scratch size at the time point when the characteristic amount changed with time reaches the nth maximum value is estimated to be n-1/2 times of the half cycle of the outer wheel circumference, and the scratch size at the time point when the characteristic amount reaches the nth minimum value is estimated to be n times of the half cycle of the outer wheel circumference, or

The scratch size at the time point when the characteristic amount changed with time reaches the nth maximum value is estimated to be n-1/2 times the distance between the rolling elements, and the scratch size at the time point when the characteristic amount reaches the nth minimum value is estimated to be n times the distance between the rolling elements.

5. The diagnostic device of claim 3, wherein

The estimating unit estimates that the inner wheel has a scratch,

estimating a scratch size at a time point when the characteristic amount changed with time reaches an initial maximum value and a time point when a difference between an nth maximum value and an n-1 st maximum value of the characteristic amount changed with time is smaller than a predetermined value as a size n-1/2 times a distance between the rolling elements, estimating a scratch size at a time point when the characteristic amount reaches the nth minimum value as a size n times the distance between the rolling elements,

when the difference between the nth maximum value and the (n-1) th maximum value of the characteristic amount that changes with time is equal to or greater than the predetermined value, it is estimated that two scratches are generated in the inner wheel.

6. The diagnostic device of claim 5, wherein

The estimating unit estimates that, when two scratches are generated in the inner wheel, the size of one of the scratches is larger than the size of the scratch at the previous estimation, and the size of the other scratch is smaller than the size of the one scratch.

7. The diagnostic device of claim 6, wherein

The estimating unit estimates the sizes of the two scratches at the time point when the characteristic amount that changes with time reaches the nth maximum value, based on a combination in which the difference between the size of the one scratch and the size of the other scratch is n times the distance between the rolling elements, and estimates the sizes of the two scratches at the time point when the sum of the size of the one scratch and the size of the other scratch becomes n times or 1/2 times the distance between the rolling elements.

8. The diagnostic device of claim 7, wherein

The estimating unit estimates a maximum value or an average value of the combinations as the sizes of the two scratches.

9. The diagnostic device of claim 7 or 8, wherein

The estimating unit estimates the size of the scratch at the time when the nth maximum value or minimum value is reached based on the temporal change in the feature amount, and estimates the sizes of the two scratches from the combination of values in a predetermined range including the estimated sizes.

10. The diagnostic device of any one of claims 1 to 8, wherein

The extraction unit normalizes the extracted feature amount using the first feature amount and the second feature amount according to the rotation speed of the bearing mechanism when data is acquired by the acquisition unit, or a predetermined value and the second feature amount, which are predetermined as values corresponding to the first feature amount, determined based on a predetermined relationship between the rotation speed of the bearing mechanism and the first feature amount in a state where no scratches are present and the second feature amount in a state where predetermined scratches are present.

11. The diagnostic device of claim 10, wherein

The extraction unit normalizes the extracted feature amount by using the first feature amount or the predetermined value as a minimum value and the second feature amount as a maximum value.

12. The diagnostic device of claim 10, wherein

The extraction unit uses, as the second feature amount, a value within a predetermined range including a maximum value of the feature amount that changes with time.

13. The diagnostic device of claim 11, wherein

The extraction unit uses, as the second feature amount, a value within a predetermined range including a maximum value of the feature amount that changes with time.

14. A diagnostic method comprising:

an acquisition unit acquires data relating to vibration corresponding to rotation of a bearing mechanism including a rolling element between an outer ring and an inner ring;

an extraction unit that extracts a feature amount from a result of frequency analysis of the data acquired by the acquisition unit;

an estimating unit that estimates the size of the scratch generated on the outer wheel or the inner wheel based on a predetermined relationship between the change in the characteristic amount and the size of the scratch generated on the outer wheel or the inner wheel, and the characteristic amount extracted by the extracting unit; and

the output unit outputs the estimation result obtained by the estimation unit.

15. A computer-readable storage medium storing a diagnostic program for causing a computer to function as:

an acquisition unit that acquires data relating to vibration corresponding to rotation of a bearing mechanism that includes a rolling element between an outer ring and an inner ring;

an extraction unit that extracts a feature amount from a result of frequency analysis of the data acquired by the acquisition unit;

an estimating unit that estimates a size of a scratch occurring in the outer wheel or the inner wheel based on a predetermined relationship between a change in the characteristic amount and a size of a scratch occurring in the outer wheel or the inner wheel and the characteristic amount extracted by the extracting unit; and

and an output unit that outputs the estimation result obtained by the estimation unit.

Images