A motor fault diagnosis method for a grinding wheel robot

Abstract

The present invention discloses a motor fault diagnosis method for a grinding wheel robot, which belongs to the field of equipment detection technology. The method includes five key steps: first, collecting vibration signals at different speeds; second, constructing an abnormal energy ratio sequence; then calculating the initial fault value; further counting the energy abnormality frequency under each speed condition; finally, compensating the initial fault value in combination with the abnormal frequency, and finally obtaining an accurate motor fault diagnosis value. The present invention collects vibration signals under three different speed conditions: high speed, medium speed and low speed, and realizes a more comprehensive and accurate motor fault assessment through energy ratio sequence construction and abnormal frequency statistics.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment detection, and particularly to a method for diagnosing motor faults of a grinding wheel robot. Background Art

[0002] In modern industrial production, as an important surface processing equipment, the stable operation of the motor of a grinding wheel robot is crucial for production efficiency and product quality. Traditional methods for diagnosing motor faults mainly rely on manual detection and empirical judgment, and there are many limitations. Existing technologies usually analyze vibration signals at a single speed or extract simple frequency-domain features to identify possible abnormalities of the motor, and it is difficult to comprehensively and accurately reflect the actual operating state of the motor.

[0003] Specifically, the existing motor fault diagnosis technologies mainly have the following problems: First, traditional methods mostly analyze vibration signals at fixed frequencies or under a single working condition, and cannot comprehensively capture the subtle changes of the motor at different speeds. Second, the existing technologies have relatively single methods for processing vibration signals, usually only focusing on amplitude or spectral features and ignoring the importance of signal energy distribution. Therefore, the existing technologies have the problem of low accuracy in diagnosing motor faults. Summary of the Invention

[0004] In view of the above deficiencies in the prior art, the method for diagnosing motor faults of a grinding wheel robot provided by the present invention solves the problem of low accuracy in diagnosing motor faults existing in the prior art.

[0005] In order to achieve the above invention object, the technical solution adopted by the present invention is: A method for diagnosing motor faults of a grinding wheel robot, comprising the following steps:

[0006] S1. Collect vibration signals of the motor of the grinding wheel robot at high speed, medium speed, and low speed respectively;

[0007] S2. Construct a high-speed abnormal energy ratio sequence, a medium-speed abnormal energy ratio sequence, and a low-speed abnormal energy ratio sequence respectively according to the energy ratios of the vibration signals at high speed, medium speed, and low speed;

[0008] S3. Calculate the initial fault value of the motor according to the high-speed abnormal energy ratio sequence, the medium-speed abnormal energy ratio sequence, and the low-speed abnormal energy ratio sequence;

[0009] S4. Count the number of elements in the high-speed abnormal energy ratio sequence, the medium-speed abnormal energy ratio sequence, and the low-speed abnormal energy ratio sequence respectively to obtain the high-speed energy abnormal frequency, the medium-speed energy abnormal frequency, and the low-speed energy abnormal frequency;

[0010] S5. Compensate the initial motor fault value based on the high-speed energy anomaly frequency, medium-speed energy anomaly frequency, and low-speed energy anomaly frequency to obtain the motor fault diagnosis value.

[0011] Further, in S1, the high speed of the motor is in the range of 3000 - 6000 rpm, the medium speed of the motor is in the range of 1500 - 3000 rpm, and the low speed of the motor is in the range of 500 - 1500 rpm, where rpm is the unit of rotational speed.

[0012] Further, S2 includes the following sub-steps:

[0013] S21. Set a sliding sequence, which slides on the vibration signals corresponding to high speed, medium speed, and low speed with the same time length respectively. The length of the sliding sequence is N, where N is a positive integer. When N is even, it advances N / 2 vibration amplitudes each time; when N is odd, it advances (N + 1) / 2 vibration amplitudes each time.

[0014] S22. After each slide, calculate the vibration energy based on the vibration amplitudes in the sliding sequence and construct a vibration energy sequence.

[0015] S23. Calculate the energy ratio of each vibration energy in the vibration energy sequence.

[0016] S24. Extract the energy ratios greater than the energy ratio threshold in the high-speed case as the high-speed abnormal energy ratios and construct a high-speed abnormal energy ratio sequence, where the energy ratio threshold is the threshold set for the energy ratios.

[0017] S25. Extract the energy ratios greater than the energy ratio threshold in the medium-speed case as the medium-speed abnormal energy ratios and construct a medium-speed abnormal energy ratio sequence.

[0018] S26. Extract the energy ratios greater than the energy ratio threshold in the low-speed case as the low-speed abnormal energy ratios and construct a low-speed abnormal energy ratio sequence.

[0019] Further, the formula for calculating the vibration energy in S22 is:

[0020] , where E i is the i-th vibration energy in the vibration energy sequence, x i,j is the j-th vibration amplitude in the sliding sequence after the i-th slide, and i and j are positive integers.

[0021] Further, the formula for calculating the energy ratio of each vibration energy in S23 is:

[0022] , where ε i is the energy ratio of the i-th vibration energy, Ei is the i-th vibration energy in the vibration energy sequence, E i-1 is the (i - 1)-th vibration energy in the vibration energy sequence, E i+1 is the (i + 1)-th vibration energy in the vibration energy sequence, M is the number of vibration energies in the vibration energy sequence, and i is a positive integer.

[0023] Furthermore, the said S3 includes the following sub-steps:

[0024] S31. Calculate the fault coefficient for the high-speed abnormal energy ratio sequence to obtain the high-speed fault coefficient;

[0025] S32. Calculate the fault coefficient for the medium-speed abnormal energy ratio sequence to obtain the medium-speed fault coefficient;

[0026] S33. Calculate the fault coefficient for the low-speed abnormal energy ratio sequence to obtain the low-speed fault coefficient;

[0027] S34. Add the high-speed fault coefficient, the medium-speed fault coefficient, and the low-speed fault coefficient to obtain the initial motor fault value.

[0028] Furthermore, the formulas for calculating the fault coefficient in S31, S32, and S33 are all:

[0029] , where γ is the fault coefficient, ε th is the energy ratio threshold, m is a positive integer, ε m is the m-th element in the sequence, L is the number of elements in the sequence, and e is the natural constant.

[0030] Furthermore, the said S5 includes the following sub-steps:

[0031] S51. Calculate the low-speed compensation coefficient according to the difference between the low-speed energy abnormal frequency and the energy abnormal frequency threshold;

[0032] S52. Calculate the medium-speed compensation coefficient according to the difference between the medium-speed energy abnormal frequency and the low-speed energy abnormal frequency;

[0033] S53. Calculate the high-speed compensation coefficient according to the difference between the high-speed energy abnormal frequency and the medium-speed energy abnormal frequency;

[0034] S54. Compensate the initial motor fault value according to the low-speed compensation coefficient, the medium-speed compensation coefficient, and the high-speed compensation coefficient to obtain the motor fault diagnosis value.

[0035] Furthermore, the formula for calculating the low-speed compensation coefficient in S51 is:

[0036] , and the formula for calculating the medium-speed compensation coefficient in S52 is:

[0037] The formula for calculating the high-speed compensation coefficient in S53 is as follows:

[0038] , where θ low is the low-speed compensation coefficient, θ middle is the medium-speed compensation coefficient, θ high is the high-speed compensation coefficient, f low is the low-speed energy anomaly frequency, f middle is the medium-speed energy anomaly frequency, f high is the high-speed energy anomaly frequency, f th is the energy anomaly frequency threshold set for the low-speed condition, C is the denominator coefficient, C > 0, and | | represents the absolute value.

[0039] Furthermore, the compensation formula in S54 is:

[0040] , where Y is the motor fault diagnosis value, y is the initial motor fault value, θ low is the low-speed compensation coefficient, θ middle is the medium-speed compensation coefficient, θ high is the high-speed compensation coefficient.

[0041] In summary, the beneficial effects of the present invention are as follows:

[0042] 1. By comprehensively collecting vibration signals under three working conditions of high speed, medium speed, and low speed and analyzing the energy ratio sequence at each speed, the present invention can capture the subtle anomalies during the operation of the motor more comprehensively and accurately. Compared with the traditional diagnosis method under a single working condition, the present invention improves the accuracy and reliability of motor fault diagnosis.

[0043] 2. The present invention extracts the energy ratio of the vibration signals at high speed, medium speed, and low speed respectively, so as to screen out the abnormal energy ratios and deeply analyze the signal energy distribution, thereby comprehensively reflecting the operating characteristics of the motor at different rotational speeds.

[0044] 3. According to the abnormal energy ratio sequences at high speed, medium speed, and low speed, the present invention calculates the initial motor fault value, and then compensates the initial motor fault value according to the high-speed, medium-speed, and low-speed energy anomaly frequencies, taking into account the energy anomaly situation and the occurrence frequency, comprehensively evaluating the motor fault situation, and improving the accuracy of fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flowchart of a motor fault diagnosis method for a grinding wheel robot. DETAILED DESCRIPTION OF THE INVENTION

[0046] The following describes the specific embodiments of the present invention to facilitate those skilled in the art of this technology to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art of this technology, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions created using the concept of the present invention are within the scope of protection.

[0047] As Figure 1 shown, a method for diagnosing motor faults of a grinding wheel robot includes the following steps:

[0048] S1. Collect vibration signals: Collect the vibration signals of the motor of the grinding wheel robot at high speed, medium speed, and low speed respectively;

[0049] S2. Construct abnormal energy ratio sequences: According to the energy ratios of the vibration signals at high speed, medium speed, and low speed, construct a high-speed abnormal energy ratio sequence, a medium-speed abnormal energy ratio sequence, and a low-speed abnormal energy ratio sequence respectively;

[0050] S3. Calculate the initial motor fault value: Calculate the initial motor fault value according to the high-speed abnormal energy ratio sequence, the medium-speed abnormal energy ratio sequence, and the low-speed abnormal energy ratio sequence;

[0051] S4. Statistically count the frequency of energy anomalies: Statistically count the number of elements in the high-speed abnormal energy ratio sequence, the medium-speed abnormal energy ratio sequence, and the low-speed abnormal energy ratio sequence respectively to obtain the high-speed energy anomaly frequency, the medium-speed energy anomaly frequency, and the low-speed energy anomaly frequency. Among them, the high-speed energy anomaly frequency is equal to the number of elements in the high-speed abnormal energy ratio sequence, the medium-speed energy anomaly frequency is equal to the number of elements in the medium-speed abnormal energy ratio sequence, and the low-speed energy anomaly frequency is equal to the number of elements in the low-speed abnormal energy ratio sequence;

[0052] S5. Compensate the initial motor fault value: Compensate the initial motor fault value according to the high-speed energy anomaly frequency, the medium-speed energy anomaly frequency, and the low-speed energy anomaly frequency to obtain the motor fault diagnosis value.

[0053] In this embodiment, the high speed of the motor in S1 is in the range of 3000 - 6000 rpm, the medium speed of the motor is in the range of 1500 - 3000 rpm, and the low speed of the motor is in the range of 500 - 1500 rpm, where rpm is the unit of rotational speed.

[0054] In this embodiment, the S2 includes the following sub-steps:

[0055] S21. Set a sliding sequence and slide it on the vibration signals corresponding to high speed, medium speed, and low speed with the same time length respectively. The length of the sliding sequence is N, where N is a positive integer. When N is an even number, move forward by N / 2 vibration amplitudes each time; when N is an odd number, move forward by (N + 1) / 2 vibration amplitudes each time.

[0056] For example: the high-speed vibration signal is [0.1, 0.2, 0.15, 0.3, 0.25, 0.4, 0.35, 0.5], N = 4 (even number). The first slide: [0.1, 0.2, 0.15, 0.3]; the second slide: [0.15, 0.3, 0.25, 0.4]; the third slide: [0.25, 0.4, 0.35, 0.5].

[0057] S22. After each slide, calculate the vibration energy according to the vibration amplitudes in the sliding sequence and construct a vibration energy sequence.

[0058] S23. Calculate the energy ratio of each vibration energy in the vibration energy sequence.

[0059] S24. Extract the energy ratios greater than the energy ratio threshold in the high-speed case as the high-speed abnormal energy ratios and construct a high-speed abnormal energy ratio sequence, where the energy ratio threshold is a threshold set for the energy ratios.

[0060] Step S24 is for the energy ratios of each vibration energy in the vibration energy sequence in the high-speed case.

[0061] S25. Extract the energy ratios greater than the energy ratio threshold in the medium-speed case as the medium-speed abnormal energy ratios and construct a medium-speed abnormal energy ratio sequence.

[0062] Step S25 is for the energy ratios of each vibration energy in the vibration energy sequence in the medium-speed case.

[0063] S26. Extract the energy ratios greater than the energy ratio threshold in the low-speed case as the low-speed abnormal energy ratios and construct a low-speed abnormal energy ratio sequence.

[0064] Step S26 is for the energy ratios of each vibration energy in the vibration energy sequence in the low-speed case.

[0065] When setting the sliding sequence in the present invention, move forward by N / 2 or (N + 1) / 2 vibration amplitudes each time. While ensuring the signal continuity, it avoids completely repeated sampling. By slightly staggering the sampling window, it can capture the subtle features that may be missed by completely overlapping windows.

[0066] In the present invention, there are three types of vibration signals: high speed, medium speed, and low speed. Therefore, there are also three vibration energy sequences.

[0067] In this embodiment, the formula for calculating the vibration energy in S22 is as follows:

[0068] , where E i is the i-th vibration energy in the vibration energy sequence, x i,j is the j-th vibration amplitude in the sliding sequence after the i-th sliding, and i and j are positive integers.

[0069] In this embodiment, the formula for calculating the energy ratio of each vibration energy in S23 is as follows:

[0070] , where ε i is the energy ratio of the i-th vibration energy, E i is the i-th vibration energy in the vibration energy sequence, E i-1 is the (i - 1)-th vibration energy in the vibration energy sequence, E i+1 is the (i + 1)-th vibration energy in the vibration energy sequence, M is the number of vibration energies in the vibration energy sequence, and i is a positive integer.

[0071] The present invention counts the vibration energy corresponding to the sliding sequence after each sliding, reflects the vibration condition of this section of the signal, and then compares each vibration energy with the mean value of the vibration energy and the neighborhood vibration energies E i-1 and E i+1 to obtain the energy ratio, reflecting the energy bulge condition.

[0072] When calculating the energy ratio of the first vibration energy in the vibration energy sequence, there is only one neighborhood vibration energy E i+1 , so only consider the ratio of each vibration energy to the mean value of the vibration energy and the neighborhood vibration energy E i+1 ; when calculating the energy ratio of the last vibration energy in the vibration energy sequence, there is only one neighborhood vibration energy E i-1 , so only consider the ratio of each vibration energy to the mean value of the vibration energy and the neighborhood vibration energy E i-1 .

[0073] In this embodiment, S3 includes the following sub-steps:

[0074] S31. Calculate the fault coefficient for the high-speed abnormal energy ratio sequence to obtain the high-speed fault coefficient;

[0075] S32. Calculate the fault coefficient for the medium-speed abnormal energy ratio sequence to obtain the medium-speed fault coefficient;

[0076] S33. Calculate the fault coefficient for the low-speed abnormal energy ratio sequence to obtain the low-speed fault coefficient;

[0077] S34. Add the high-speed fault coefficient, medium-speed fault coefficient, and low-speed fault coefficient to obtain the initial motor fault value.

[0078] In this embodiment, the formulas for calculating the fault coefficient in S31, S32, and S33 are all:

[0079] , where γ is the fault coefficient, ε th is the energy ratio threshold, m is a positive integer, and ε m is the m-th element in the sequence, L is the number of elements in the sequence, and e is the natural constant.

[0080] The present invention calculates the fault coefficient based on the difference between each abnormal energy ratio value in the sequence and the energy ratio threshold. When the difference is larger, the abnormal situation of the abnormal energy ratio value is greater, and it is more significant compared to the energy of other parts. The present invention also sets an enhancement coefficient to further enhance the abnormal situation, so that the part with more significant energy has a more significant fault coefficient.

[0081] In this embodiment, the energy ratio values obtained through step S23 at high speed are used to construct a high-speed energy ratio value sequence, the energy ratio values obtained through step S23 at medium speed are used to construct a medium-speed energy ratio value sequence, and the energy ratio values obtained through step S23 at low speed are used to construct a low-speed energy ratio value sequence; through step S24, that is, screening the high-speed abnormal energy ratio values from the high-speed energy ratio value sequence to construct a high-speed abnormal energy ratio value sequence, through step S25, that is, screening the medium-speed abnormal energy ratio values from the medium-speed energy ratio value sequence to construct a medium-speed abnormal energy ratio value sequence, and through step S26, that is, screening the low-speed abnormal energy ratio values from the low-speed energy ratio value sequence to construct a low-speed abnormal energy ratio value sequence.

[0082] In this embodiment, the energy ratio threshold is set to 1.5 times the average energy ratio value of the sequence. For the high-speed situation, the energy ratio threshold is 1.5 times the average energy ratio value of the high-speed energy ratio value sequence, named the high-speed energy ratio threshold; for the medium-speed situation, the energy ratio threshold is 1.5 times the average energy ratio value of the medium-speed energy ratio value sequence, named the medium-speed energy ratio threshold; for the low-speed situation, the energy ratio threshold is 1.5 times the average energy ratio value of the low-speed energy ratio value sequence, named the low-speed energy ratio threshold. Therefore, in the fault coefficient formula, when the sequence is the high-speed abnormal energy ratio value sequence, the energy ratio threshold is the high-speed energy ratio threshold; when the sequence is the medium-speed abnormal energy ratio value sequence, the energy ratio threshold is the medium-speed energy ratio threshold, and when the sequence is the low-speed abnormal energy ratio value sequence, the energy ratio threshold is the low-speed energy ratio threshold. At the same time, the energy ratio threshold in S24 is the high-speed energy ratio threshold, the energy ratio threshold in S25 is the medium-speed energy ratio threshold, and the energy ratio threshold in S26 is the low-speed energy ratio threshold.

[0083] In this embodiment, energy ratio thresholds of the same magnitude can also be set for three cases of high speed, medium speed, and low speed.

[0084] In this embodiment, S5 includes the following sub-steps:

[0085] S51. Calculate a low-speed compensation coefficient according to the difference between the low-speed energy anomaly frequency and the energy anomaly frequency threshold;

[0086] S52. Calculate a medium-speed compensation coefficient according to the difference between the medium-speed energy anomaly frequency and the low-speed energy anomaly frequency;

[0087] S53. Calculate a high-speed compensation coefficient according to the difference between the high-speed energy anomaly frequency and the medium-speed energy anomaly frequency;

[0088] S54. Compensate the initial motor fault value according to the low-speed compensation coefficient, the medium-speed compensation coefficient, and the high-speed compensation coefficient to obtain a motor fault diagnosis value.

[0089] In this embodiment, the energy anomaly frequency threshold is the threshold of the energy anomaly frequency set for the low-speed case, and this value can be specifically set according to experience or requirements. When the fault detection accuracy requirement is high, the energy anomaly frequency threshold can be set to be less than or equal to 5. When the fault detection accuracy requirement is low, the energy anomaly frequency threshold can be set to be greater than 5.

[0090] In this embodiment, the formula for calculating the low-speed compensation coefficient in S51 is:

[0091] , and the formula for calculating the medium-speed compensation coefficient in S52 is:

[0092] , and the formula for calculating the high-speed compensation coefficient in S53 is:

[0093] , where θ low is the low-speed compensation coefficient, θ middle is the medium-speed compensation coefficient, θ high is the high-speed compensation coefficient, f low is the low-speed energy anomaly frequency, f middle is the medium-speed energy anomaly frequency, f high is the high-speed energy anomaly frequency, f th is the energy anomaly frequency threshold set for the low-speed case, C is the denominator coefficient, C is greater than 0, and | | is the absolute value.

[0094] In this embodiment, the compensation formula in S54 is:

[0095] , where Y is the motor fault diagnosis value, y is the initial motor fault value, θlow is the low-speed compensation coefficient, θ middle is the medium-speed compensation coefficient, θ high is the high-speed compensation coefficient.

[0096] The vibration signal itself will be affected by other components. Therefore, there are some abnormal energy situations that are affected by other noises. In order to eliminate the influence of other vibrations and further improve the fault diagnosis accuracy, the present invention first calculates the low-speed compensation coefficient according to the difference between the low-speed energy abnormal frequency and the energy abnormal frequency threshold. When the low-speed energy abnormal frequency is less than the energy abnormal frequency threshold, the low-speed compensation coefficient is negative and the motor fault probability is low. When the low-speed energy abnormal frequency is greater than the energy abnormal frequency threshold, the low-speed compensation coefficient is positive and the motor fault probability is high. At medium speed and high speed, when it belongs to the motor's own fault, as the speed increases, the energy abnormal frequency should be greater. If the energy abnormal frequency decreases as the speed increases, the motor fault probability is small, which belongs to the vibration abnormality caused by the influence of other vibrations. Therefore, the present invention calculates the medium-speed compensation coefficient according to the difference between the medium-speed energy abnormal frequency and the low-speed energy abnormal frequency, and calculates the high-speed compensation coefficient according to the difference between the high-speed energy abnormal frequency and the medium-speed energy abnormal frequency, reflecting whether the number of abnormal energy ratios increases with the increase of the rotational speed under different speed conditions. If it belongs to the motor's fault, the number of abnormal energy ratios should increase with the increase of the rotational speed.

[0097] The present invention comprehensively combines the low-speed compensation coefficient, the medium-speed compensation coefficient and the high-speed compensation coefficient to compensate the initial motor fault value and improve the accuracy of motor fault diagnosis.

[0098] By comprehensively collecting vibration signals under three working conditions of high speed, medium speed and low speed and analyzing the energy ratio sequence at each speed, the present invention can capture the subtle abnormalities in the motor operation process more comprehensively and accurately. Compared with the traditional diagnosis method under a single working condition, the present invention improves the accuracy and reliability of motor fault diagnosis.

[0099] The present invention extracts the energy ratios of the vibration signals at high speed, medium speed and low speed respectively, so as to screen out the abnormal energy ratios and deeply analyze the signal energy distribution, so as to comprehensively reflect the operation characteristics of the motor at different rotational speeds.

[0100] The present invention calculates the initial motor fault value according to the abnormal energy ratio sequences at high speed, medium speed and low speed, and then compensates the initial motor fault value according to the high-speed, medium-speed and low-speed energy abnormal frequencies, considering the abnormal energy situation and the occurrence frequency, comprehensively evaluating the motor fault situation and improving the accuracy of fault diagnosis.

[0101] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A motor fault diagnosis method for a grinding wheel robot, characterized in that: The following steps are involved: S1, respectively collecting vibration signals of the motor of the grinding wheel robot at high speed, medium speed and low speed; S2. According to the energy ratios of the vibration signals at high speed, medium speed and low speed, a high-speed abnormal energy ratio sequence, a medium-speed abnormal energy ratio sequence and a low-speed abnormal energy ratio sequence are constructed respectively, wherein the formula of the energy ratio is: , where ε i is the energy ratio of the i-th vibration energy, E i is the i-th vibration energy in the vibration energy sequence, E i-1 is the i-1th vibration energy in the vibration energy sequence, E i+1 is the i+1th vibration energy in the vibration energy sequence, M is the number of vibration energies in the vibration energy sequence, and i is a positive integer; S3, calculating the initial fault value of the motor according to the high-speed abnormal energy ratio sequence, the medium-speed abnormal energy ratio sequence and the low-speed abnormal energy ratio sequence; The S3 comprises the following sub-steps: S31, calculating the failure coefficient for the high-speed abnormal energy ratio sequence to obtain the high-speed failure coefficient; S32, calculating the fault coefficient for the medium-speed abnormal energy ratio sequence to obtain the medium-speed fault coefficient; S33, calculating the fault coefficient for the low-speed abnormal energy ratio sequence to obtain the low-speed fault coefficient; S34, adding the high-speed fault coefficient, the medium-speed fault coefficient and the low-speed fault coefficient to obtain an initial fault value of the motor; The formulas for calculating the failure coefficient are: , where γ is the failure coefficient, ε th is the energy ratio threshold, m is a positive integer, ε m is the mth element in the sequence, L is the number of elements in the sequence, and e is a natural constant; S4, respectively counting the number of elements in the high-speed abnormal energy ratio sequence, the medium-speed abnormal energy ratio sequence, and the low-speed abnormal energy ratio sequence to obtain the high-speed energy abnormal frequency, the medium-speed energy abnormal frequency, and the low-speed energy abnormal frequency; S5. According to the high-speed energy abnormality frequency, the medium-speed energy abnormality frequency and the low-speed energy abnormality frequency, the initial fault value of the motor is compensated to obtain a motor fault diagnosis value.

2. The motor fault diagnosis method of the grinding wheel robot according to claim 1, characterized in that: In the S1, the high speed of the motor is in the range of 3000-6000 rpm, the medium speed of the motor is in the range of 1500-3000 rpm, and the low speed of the motor is in the range of 500-1500 rpm, wherein rpm is the unit of rotation speed.

3. The motor fault diagnosis method of the grinding wheel robot according to claim 1, characterized in that: The S2 comprises the following sub-steps: S21, set a sliding sequence, slide on the vibration signals corresponding to high speed, medium speed and low speed of the same time length respectively, the length of the sliding sequence is N, N is a positive integer, when N is an even number, each time advances N / 2 vibration amplitudes, when N is an odd number, each time advances (N+1) / 2 vibration amplitudes; S22, after each sliding, calculating the vibration energy according to the vibration amplitude in the sliding sequence, and constructing a vibration energy sequence; S23, calculating the energy ratio of each vibration energy in the vibration energy sequence; S24, extracting energy ratio values ​​greater than an energy ratio threshold value under high-speed conditions as high-speed abnormal energy ratio values, and constructing a high-speed abnormal energy ratio value sequence, wherein the energy ratio threshold value is a threshold value set for the energy ratio value; S25, extracting the energy ratio value greater than the energy ratio threshold value under the medium speed condition as the medium speed abnormal energy ratio value, and constructing the medium speed abnormal energy ratio value sequence; S26. Extract the energy ratio value greater than the energy ratio threshold value in the low-speed condition as the low-speed abnormal energy ratio value, and construct a low-speed abnormal energy ratio value sequence.

4. The motor fault diagnosis method of the grinding wheel robot according to claim 3 is characterized in that: The formula for calculating the vibration energy in S22 is: , where E i is the i-th vibration energy in the vibration energy sequence, x i,j is the jth vibration amplitude in the sliding sequence after the i-th sliding, where i and j are positive integers.

5. The motor fault diagnosis method of the grinding wheel robot according to claim 1, characterized in that: The S5 comprises the following sub-steps: S51, calculating a low-speed compensation coefficient according to a difference between the low-speed energy abnormality frequency and the energy abnormality frequency threshold, wherein the energy abnormality frequency threshold is a threshold of the energy abnormality frequency set for the low-speed situation; S52, calculating a medium-speed compensation coefficient according to a difference between the medium-speed energy abnormality frequency and the low-speed energy abnormality frequency; S53, calculating a high-speed compensation coefficient according to a difference between the high-speed energy abnormal frequency and the medium-speed energy abnormal frequency; S54. Compensate the initial fault value of the motor according to the low-speed compensation coefficient, the medium-speed compensation coefficient and the high-speed compensation coefficient to obtain a motor fault diagnosis value.

6. The motor fault diagnosis method of the grinding wheel robot according to claim 5, characterized in that: The formula for calculating the low speed compensation coefficient in S51 is: , the formula for calculating the medium speed compensation coefficient in S52 is: , the formula for calculating the high speed compensation coefficient in S53 is: , where θ low is the low speed compensation coefficient, θ middle is the medium speed compensation coefficient, θ high is the high speed compensation coefficient, f low is the frequency of low-speed energy anomaly, f middle is the frequency of medium-speed energy anomaly, f high is the high-speed energy abnormal frequency, f th is the energy abnormal frequency threshold set for low-speed conditions, C is the denominator coefficient, C is greater than 0, and | | is the absolute value.

7. The motor fault diagnosis method of the grinding wheel robot according to claim 5, characterized in that: The compensation formula in S54 is: , where Y is the motor fault diagnosis value, y is the motor initial fault value, θ low is the low speed compensation coefficient, θ middle is the medium speed compensation coefficient, θ high is the high speed compensation coefficient.

Images