Operation fault monitoring method, system and equipment suitable for permanent magnet synchronous motor

Abstract

The invention relates to the technical field of motor fault monitoring, in particular to an operation fault monitoring method, system and device suitable for a permanent magnet synchronous motor, and the method comprises the steps: obtaining the operation parameters of a current statistical period and a previous statistical period of the permanent magnet synchronous motor, the harmonic degree of each operation parameter is obtained according to the number of different subharmonics in each operation parameter and the abnormal condition of each subharmonic, and the correlation degree of the operation parameters of the two statistical periods is obtained according to the harmonic degrees of the two statistical periods and the change of the three-phase current unbalance degree; according to the magnitude relationship between the correlation degree and the change degree of the operation task, the reference weight of the harmonic degree is determined, and the fault monitoring model is trained in combination with other features, so that the accuracy of the motor fault monitoring model can be improved, and the accuracy of the operation fault detection result of the motor is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of motor fault monitoring, and particularly relates to an operating fault monitoring method, system and device applicable to a permanent magnet synchronous motor. Background Art

[0002] Due to advantages such as high efficiency and high power density, permanent magnet synchronous motors are widely used in many fields such as industry and new energy vehicles. Since operating faults such as "permanent magnet demagnetization", "bearing damage", and "winding short circuit" may cause equipment failure or even safety accidents during its operation, real-time monitoring technology is crucial. Currently, the industry mainly uses intelligent diagnostic algorithms such as online parameter identification, multi-sensor fusion monitoring, and neural networks for fault monitoring. Among them, the fault monitoring method using neural networks is as follows: a training set is constructed according to the operating characteristics of the permanent magnet synchronous motor, and a fault monitoring model is obtained through model training based on the training set, so as to realize the operating fault monitoring of the permanent magnet synchronous motor according to the fault monitoring model. Among them, when obtaining the training set, the reference weights are usually not set for all the characteristics of the permanent magnet synchronous motor, that is, the reference weights of all the characteristics of the permanent magnet synchronous motor are 1. This fixed reference weight setting method has the following defects: when identifying the operating faults of the permanent magnet synchronous motor, since harmonics are relatively common features in the fault manifestations, the more harmonics appear and the greater the degree of fluctuation, the greater the possibility that the corresponding permanent magnet synchronous motor has a fault. However, harmonics may also be caused by other reasons in addition to abnormal operation. For example, if the operation task of the permanent magnet synchronous motor changes, if this situation is not considered, it will result in the lack of pertinence of each characteristic in the training set, thereby reducing the accuracy of the fault monitoring model and easily causing misjudgment of the operating fault detection results of the permanent magnet synchronous motor. Summary of the Invention

[0003] In order to solve the technical problem of low training accuracy of the existing fault monitoring model, the purpose of the present invention is to provide an operating fault monitoring method, system and device applicable to a permanent magnet synchronous motor. The specific technical solutions adopted are as follows: In the first aspect of the present invention, an operating fault monitoring method applicable to a permanent magnet synchronous motor is provided, including: Obtaining the operating parameters of the permanent magnet synchronous motor in the current statistical period and the previous statistical period, where the operating parameters at least include three-phase current and three-phase voltage; Obtaining the harmonic degree of each operating parameter from the number of different-order harmonics and the abnormal conditions of each order harmonic in each operating parameter; Obtaining the correlation degree of the operating parameters in the current statistical period and the previous statistical period according to the change in the harmonic degree between the current statistical period and the previous statistical period, and the change in the unbalance degree of the three-phase current; Determine the reference weight of the harmonic degree according to the magnitude relationship between the correlation degree corresponding to the current statistical period and the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period; Train a permanent magnet synchronous motor fault monitoring model based on the harmonic degree under the addition of the reference weight and other characteristics of the permanent magnet synchronous motor.

[0004] In an exemplary embodiment, the process of obtaining the harmonic degree of the operating parameters includes: Obtain the fundamental wave and data groups of each harmonic of the operating parameters, where the data groups are composed of amplitude and frequency; Respectively obtain the Euclidean distances between the data groups of the fundamental wave and each harmonic to obtain the comprehensive Euclidean distance; Fuse the comprehensive Euclidean distance and the number of different harmonics in the operating parameters to obtain the harmonic degree of the operating parameters, and the harmonic degree is proportional to both the comprehensive Euclidean distance and the number of different harmonics in the operating parameters.

[0005] In an exemplary embodiment, the process of obtaining the unbalance degree of the three-phase current includes: Obtain the difference between any two-phase currents, where the difference includes the DTW distance and the mean difference of the fluctuation curves of any two-phase currents; Fuse the differences between all any two-phase currents to obtain the unbalance degree of the three-phase current.

[0006] In an exemplary embodiment, the change in the harmonic degree between the current statistical period and the previous statistical period is the absolute value of the comprehensive difference between the harmonic degrees of the current statistical period and the previous statistical period, and the absolute value of the comprehensive difference is obtained by fusing the absolute values of the differences in the harmonic degrees of all operating parameters between the current statistical period and the previous statistical period; The change in the unbalance degree of the three-phase current is the absolute value of the difference between the unbalance degrees of the three-phase current in the current statistical period and the previous statistical period.

[0007] In an exemplary embodiment, the process of obtaining the correlation degree includes: According to the absolute value of the comprehensive difference in the harmonic degrees between the current statistical period and the previous statistical period, and the absolute value of the difference in the unbalance degree of the three-phase current, obtain the correlation degree of the operating parameters between the current statistical period and the previous statistical period; the correlation degree is inversely proportional to both the absolute value of the comprehensive difference in the harmonic degrees and the absolute value of the difference in the unbalance degree of the three-phase current.

[0008] In an exemplary embodiment, the degree of change of the operation tasks of the permanent magnet synchronous motor in the current statistical period is the proportion of the types of the operation tasks of the permanent magnet synchronous motor in the current statistical period, and the proportion is the ratio of the actual number of types of the operation tasks of the permanent magnet synchronous motor in the current statistical period to the total number of types of the operation tasks of the permanent magnet synchronous motor.

[0009] In an exemplary embodiment, the process of obtaining the reference weight of the harmonic degree includes: If the correlation degree corresponding to the current statistical period is greater than the degree of change of the operation tasks of the permanent magnet synchronous motor in the current statistical period, the reference weight of the harmonic degree is less than the first preset weight threshold, and the gap between the correlation degree corresponding to the current statistical period and the degree of change of the operation tasks of the permanent magnet synchronous motor in the current statistical period is inversely proportional to the reference weight of the harmonic degree; If the correlation degree corresponding to the current statistical period is less than the degree of change of the operation tasks of the permanent magnet synchronous motor in the current statistical period, the reference weight of the harmonic degree is greater than the second preset weight threshold, and the gap between the correlation degree corresponding to the current statistical period and the degree of change of the operation tasks of the permanent magnet synchronous motor in the current statistical period is directly proportional to the reference weight of the harmonic degree; the first preset weight threshold is less than or equal to the second preset weight threshold.

[0010] In an exemplary embodiment, the reference weight of other characteristics of the permanent magnet synchronous motor is equal to the difference between the value 1 and the reference weight of the harmonic degree; Training the permanent magnet synchronous motor fault monitoring model based on the harmonic degree under the reference weight addition and other characteristics of the permanent magnet synchronous motor includes: Obtain the respective characteristics of the permanent magnet synchronous motor in multiple statistical periods under the corresponding reference weight addition, and train the permanent magnet synchronous motor fault monitoring model with this.

[0011] In a second aspect of the present invention, there is provided an operation fault monitoring system applicable to a permanent magnet synchronous motor, including: a memory and a processor; the memory is connected to the processor; the memory is used to store program instructions; the processor is used to implement the above-mentioned operation fault monitoring method applicable to the permanent magnet synchronous motor when the program instructions are executed.

[0012] In a third aspect of the present invention, there is provided an operation fault monitoring device applicable to a permanent magnet synchronous motor, including a computer-readable storage medium, and the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned operation fault monitoring method embodiment applicable to the permanent magnet synchronous motor are implemented.

[0013] The present invention has the following beneficial effects: According to the harmonic degrees of the operating parameters of the permanent magnet synchronous motor and the change of the unbalance degree of the three-phase current in two adjacent statistical periods, the correlation degree of the operating parameters in these two statistical periods is obtained, and the magnitude relationship with the change degree of the operation task of the permanent magnet synchronous motor is determined, so as to determine the reference weight of the harmonic degree, ensuring that the reference weight of the harmonic degree is closely related to the change degree of the operation task of the permanent magnet synchronous motor and is not fixed. Considering that harmonics may be caused by other reasons in addition to abnormal operation, the features in the training set are targeted, thereby improving the accuracy of the permanent magnet synchronous motor fault monitoring model and the accuracy of the operation fault detection result of the permanent magnet synchronous motor. Description of the Drawings

[0014] Figure 1 is a flowchart of the steps of a method for monitoring the operation fault of a permanent magnet synchronous motor provided by an embodiment of the present invention; Figure 2 is a flowchart for obtaining the unbalance degree of the three-phase current provided by an embodiment of the present invention; Figure 3 is a flowchart for obtaining the harmonic degree of the operating parameters provided by an embodiment of the present invention. Detailed Embodiment

[0015] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following describes in detail the specific embodiments, structures, features and effects of the present invention in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. All data information collected in this application is obtained with full consent and authorization.

[0017] This embodiment provides a method for monitoring the operation fault of a permanent magnet synchronous motor, as Figure 1 shown, including the following steps: Step S1: Obtain the operating parameters of the permanent magnet synchronous motor in the current statistical period and the previous statistical period, and the operating parameters at least include three-phase current and three-phase voltage; Step S2: Obtain the harmonic degree of each operating parameter from the number of different harmonics in each operating parameter and the abnormal conditions of each harmonic; Step S3: Obtain the correlation degree of the operating parameters in the current statistical period and the previous statistical period based on the change in the harmonic degree between the current statistical period and the previous statistical period, and the change in the unbalance degree of the three-phase current. Step S4: Determine the reference weight of the harmonic degree according to the magnitude relationship between the correlation degree corresponding to the current statistical period and the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period. Step S5: Train the fault monitoring model of the permanent magnet synchronous motor based on the harmonic degree under the addition of the reference weight and other characteristics of the permanent magnet synchronous motor.

[0018] The following specifically describes each step in combination with the accompanying drawings.

[0019] Step S1: Obtain the operating parameters of the permanent magnet synchronous motor in the current statistical period and the previous statistical period. The operating parameters at least include three-phase current and three-phase voltage.

[0020] Set relevant data acquisition devices on the permanent magnet synchronous motor for collecting the required operating parameters. The operating parameters at least include the three-phase current and three-phase voltage of the permanent magnet synchronous motor. Then, the data acquisition devices at least include a three-phase current sensor and a three-phase voltage sensor. The three-phase current sensor and the three-phase voltage sensor are arranged on the three-phase windings of the permanent magnet synchronous motor to obtain the three-phase current and three-phase voltage respectively. In an exemplary embodiment, in addition to the three-phase current and three-phase voltage, the operating parameters also include the rotational speed. Correspondingly, a rotational speed sensor is installed on the rotating shaft of the permanent magnet synchronous motor to collect the rotational speed signal of the permanent magnet synchronous motor.

[0021] In an exemplary embodiment, a temperature sensor is also fixed on the permanent magnet synchronous motor to collect the temperature signal of the permanent magnet synchronous motor. The temperature sensor can be arranged near the stator winding or the permanent magnet.

[0022] It should be understood that each data acquisition device is signal-connected to a data acquisition card for converting the analog signal into a digital signal. Moreover, the sampling frequency of each data acquisition device is set according to actual needs.

[0023] In this embodiment, the permanent magnet synchronous motor is analyzed with the statistical period as the time unit. During the operation of the permanent magnet synchronous motor, every time a statistical period passes in time, the statistical period is included in the analysis. The duration of each statistical period is the same, and the specific duration is set according to actual needs, such as half an hour. Therefore, the statistical period includes multiple sampling moments, and each operating parameter of the permanent magnet synchronous motor is obtained once at each sampling moment, so as to obtain the operating parameter sequence corresponding to each operating parameter of the statistical period.

[0024] For the sake of convenience of explanation, it is assumed that two adjacent statistical periods are the current statistical period and the previous statistical period of the current statistical period respectively. By analyzing the operating parameters of the permanent magnet synchronous motor in the current statistical period and the previous statistical period, the reference weight of the relevant characteristics of the current statistical period of the permanent magnet synchronous motor is obtained.

[0025] Therefore, the operating parameters of the permanent magnet synchronous motor in the current statistical period and the previous statistical period are obtained. The operating parameters of each statistical period are essentially the operating parameter fluctuation curves in the time domain.

[0026] Step S2: Obtain the harmonic degree of each operating parameter from the number of different harmonics and the abnormal conditions of each harmonic in each operating parameter.

[0027] When faults such as winding short circuit and open phase occur, the three-phase current of the permanent magnet synchronous motor will show obvious imbalance. The existing method for obtaining the imbalance degree of the three-phase current is determined by dividing the difference between the maximum and minimum currents by the average value. However, when the waveform of a certain phase current has a time shift due to load changes, the imbalance degree calculated by the existing method will deviate. Therefore, it is also necessary to analyze the difference between the currents of different phases in the three-phase current to determine the imbalance degree of the three-phase current. The more serious the difference situation is, the more serious the imbalance of the three-phase current is. In order to improve the accuracy of the imbalance degree of the three-phase current, as Figure 2 shown, a specific obtaining process of the imbalance degree of the three-phase current is given as follows: Step S2-1: Obtain the difference between any two-phase currents, and the difference includes the DTW distance and the mean difference of the fluctuation curves of any two-phase currents.

[0028] For any statistical period, obtain the fluctuation curves of any two-phase currents in this statistical period, that is, the current fluctuation curves of any two-phase currents, and then obtain the DTW (Dynamic Time Warping) distance between the current fluctuation curves of these two-phase currents. The DTW distance characterizes the difference between the current fluctuation curves of these two-phase currents. The larger the DTW distance is, the greater the difference between the current fluctuation curves of these two-phase currents is.

[0029] Moreover, obtain the current mean values of the current fluctuation curves of each of these two-phase currents, so as to obtain two current mean values, and then calculate the difference between these two current mean values. Specifically, calculate the absolute value of the difference between these two current mean values as the mean difference of the current fluctuation curves of these two-phase currents. The mean difference characterizes the difference between the current fluctuation curves of these two-phase currents. The larger the mean difference is, the greater the difference between the current fluctuation curves of these two-phase currents is.

[0030] Based on the DTW distance and the mean difference between these two-phase currents, the difference between these two-phase currents is obtained. In an exemplary embodiment, the product of the DTW distance and the mean difference between these two-phase currents is calculated as the difference between these two-phase currents. Thus, the difference between any two-phase currents among the three-phase currents is obtained.

[0031] Step S2-2: Fuse the differences between any two-phase currents to obtain the unbalance degree of the three-phase currents.

[0032] Fuse the differences between any two-phase currents to obtain the unbalance degree of the three-phase currents. In an exemplary embodiment, a specific quantization method for the unbalance degree of the three-phase currents is given as follows: ; where, represents the unbalance degree of the three-phase currents, represents the current fluctuation curve of the i-th phase current among the three-phase currents, represents the current fluctuation curve of the j-th phase current among the other two-phase currents except the i-th phase current, represents the current fluctuation curve and the current fluctuation curve is the DTW distance, represents the difference between the current mean value of the current fluctuation curve and the current mean value of the current fluctuation curve is the difference. represents the normalization function, and the normalization method here can be the sigmoid function.

[0033] Since when a permanent magnet synchronous motor has an operating fault, certain harmonics will appear in the operating parameters. Taking the three-phase current and the three-phase voltage as an example, in a permanent magnet synchronous motor, the normal waveforms of the three-phase current and the three-phase voltage usually refer to the current and voltage waveforms that conform to the standard sine law. In an ideal state, the three-phase voltage and three-phase current waveforms are sine waves with a fixed frequency, a stable amplitude, and a phase difference of 120°. The harmonics of the three-phase voltage and three-phase current refer to the sine wave components whose frequencies are integer multiples of the fundamental frequency. Among them, according to different fault types, the degree of harmonics is also divided into 3rd harmonic, 5th harmonic, 7th harmonic, etc. (where the 3rd harmonic means: its frequency is 3 times the fundamental frequency. If the fundamental frequency is 50 Hz, then the frequency of the 3rd harmonic is 150 Hz).

[0034] For any operating parameter, obtain its harmonics. Taking three-phase current as an example, then, for any one of the three-phase currents, convert the current fluctuation curve of this phase current from the time domain to the frequency domain, then obtain the spectral lines corresponding to each harmonic on the frequency-domain graph, and finally determine the amplitude and frequency of each harmonic according to the height and position of the spectral lines corresponding to each harmonic. Thus, the amplitude and frequency of each harmonic of this phase current, as well as the number of different harmonics, are obtained. The currents of the other two phases are processed in this way to obtain the corresponding amplitude and frequency of each harmonic, as well as the number of different harmonics. When the operating parameter is three-phase voltage, according to the processing method of the above three-phase current, the amplitude and frequency of each harmonic corresponding to each phase voltage, as well as the number of different harmonics, are obtained. When the operating parameter is rotational speed, it is also processed according to the above three-phase current processing method to obtain the amplitude and frequency of each harmonic corresponding to the rotational speed, as well as the number of different harmonics.

[0035] When the number of different harmonics is smaller, and at the same time the amplitude and frequency corresponding to different harmonics are closer to the amplitude and frequency of the fundamental wave, it indicates that the harmonic degree of the operating parameter is lower. In an exemplary embodiment, as Figure 3 shown, a specific obtaining process of the harmonic degree of the operating parameter is given as follows: Step S2-3: Obtain the data groups of the fundamental wave and each harmonic of the operating parameter, and the data group is composed of amplitude and frequency.

[0036] For any operating parameter, obtain the data group of the fundamental wave of this operating parameter, and the data group of the fundamental wave is composed of the amplitude and frequency of the fundamental wave, and obtain the data groups of each harmonic of this operating parameter, and the data group is composed of the amplitude and frequency of the corresponding harmonic.

[0037] For three-phase current and three-phase voltage, it is necessary to separately obtain the data groups of the fundamental wave and each harmonic corresponding to each phase current, and the data groups of the fundamental wave and each harmonic corresponding to each phase voltage.

[0038] Step S2-4: Respectively obtain the Euclidean distances between the data group of the fundamental wave and the data groups of each harmonic to obtain the comprehensive Euclidean distance.

[0039] Obtain the Euclidean distances between the data group of the fundamental wave of this operating parameter and the data groups of each harmonic of this operating parameter, so as to obtain the Euclidean distances corresponding to each harmonic, and then fuse the Euclidean distances corresponding to each harmonic of this operating parameter to obtain the comprehensive Euclidean distance of this operating parameter.

[0040] For three-phase current and three-phase voltage, it is necessary to separately obtain the comprehensive Euclidean distances corresponding to each phase current, and the comprehensive Euclidean distances corresponding to each phase voltage.

[0041] Step S2-5: Integrate the comprehensive Euclidean distance and the number of different harmonics in the operating parameters to obtain the harmonic degree of the operating parameters.

[0042] The comprehensive Euclidean distance characterizes the difference between the fundamental wave and each harmonic of the operating parameters. The larger the comprehensive Euclidean distance, the greater the difference, and the greater the harmonic degree of the operating parameters. Therefore, the harmonic degree is proportional to the comprehensive Euclidean distance; the more the number of different harmonics in the operating parameters, the greater the harmonic degree. Therefore, the harmonic degree is proportional to the number of different harmonics in the operating parameters.

[0043] Taking the harmonic degree of any one phase current of the three-phase current as an example, a specific quantification method is given as follows: ; Among them, represents the harmonic degree of the k-th phase current in the three-phase current, represents the number of different harmonics that appear in the current fluctuation curve of the k-th phase current, represents the data group composed of the amplitude and frequency of the r-th harmonic in the current fluctuation curve of the k-th phase current, represents the data group composed of the amplitude and frequency of the fundamental wave in the current fluctuation curve of the k-th phase current, represents and the Euclidean distance of. represents the normalization function, and the normalization method here can be the sigmoid function.

[0044] Using the above formula, the harmonic degree of each phase current can be obtained. Similarly, the harmonic degree of each phase voltage and the harmonic degree of the rotational speed can be obtained.

[0045] Step S3: According to the change in the harmonic degree between the current statistical period and the previous statistical period, and the change in the unbalance degree of the three-phase current, obtain the correlation degree of the operating parameters between the current statistical period and the previous statistical period.

[0046] In the actual operation of the permanent magnet synchronous motor, due to the change of the operation task of the permanent magnet synchronous motor, the characteristic performance of the operating parameters is different in different statistical periods. And due to the change of the operation task, the number of harmonics of the operating parameters will change, usually increase. If these situations are not considered, it will lead to abnormal detection results, and then lead to misjudgment or missed judgment.

[0047] By adopting step S2, the harmonic degree of each operating parameter in each statistical period and the unbalance degree of the three-phase current in each statistical period can be obtained. Then, according to the change in the harmonic degree between the current statistical period and the previous statistical period and the change in the unbalance degree of the three-phase current, the correlation degree of the operating parameters between the current statistical period and the previous statistical period can be obtained.

[0048] For the three-phase current, step S2 obtains the harmonic degree of each phase current. Then, calculate the average value of the harmonic degrees of each phase current within the statistical period as the harmonic degree of the three-phase current in the statistical period. Similarly, for the three-phase voltage, step S2 obtains the harmonic degree of each phase voltage. Then, calculate the average value of the harmonic degrees of each phase voltage within the statistical period as the harmonic degree of the three-phase voltage in the statistical period.

[0049] In an exemplary embodiment, the process of obtaining the change in the harmonic degree between the current statistical period and the previous statistical period is as follows: Obtain the absolute value of the difference in the harmonic degrees of each operating parameter between the current statistical period and the previous statistical period, and then fuse the absolute values of the differences in the harmonic degrees of each operating parameter between the current statistical period and the previous statistical period to obtain the comprehensive absolute value of the difference in the harmonic degree between the current statistical period and the previous statistical period.

[0050] The following gives a specific quantification method for the change in the harmonic degree between the current statistical period and the previous statistical period: ; where represents the degree of change in the harmonic degree of the (M + 1)-th statistical period compared to the M-th statistical period, that is, the change in the harmonic degree between the (M + 1)-th statistical period and the M-th statistical period. represents the harmonic degree of the e-th operating parameter in the (M + 1)-th statistical period. represents the harmonic degree of the e-th operating parameter in the M-th statistical period. The reason why e is equal to 3 is that there are three operating parameters, namely three-phase current, three-phase voltage, and rotational speed.

[0051] represents the normalization function, and the normalization method here can be the sigmoid function.

[0052] To obtain the change in the unbalance degree of the three-phase current between the current statistical period and the previous statistical period, it is: the absolute value of the difference in the unbalance degree of the three-phase current between the current statistical period and the previous statistical period. The calculation formula is: ; where represents the change in the unbalance degree of the three-phase current between the (M + 1)-th statistical period and the M-th statistical period. Indicates the unbalance degree of the three-phase current in the (M + 1)-th statistical period. Indicates the unbalance degree of the three-phase current in the M-th statistical period.

[0053] Since the change in the harmonic level may exist independently of the change in the unbalance degree of the three-phase current, that is, when the harmonic level changes in the current statistical period compared with the previous statistical period, the unbalance degree of the three-phase current may not change. In this case, the change in the harmonic level may be affected by factors such as harmonic injection from the power supply side and non-linear loads. Therefore, by comprehensively considering the change in the harmonic level and the change in the unbalance degree of the three-phase current between the current statistical period and the previous statistical period, the correlation degree of the operating parameters between the current statistical period and the previous statistical period is obtained.

[0054] In an exemplary embodiment, the process of obtaining the correlation degree is specifically as follows: Based on the absolute value of the comprehensive difference in the harmonic level between the current statistical period and the previous statistical period, and the absolute value of the difference in the unbalance degree of the three-phase current between the current statistical period and the previous statistical period, the correlation degree of the operating parameters between the current statistical period and the previous statistical period is obtained. The larger the absolute value of the comprehensive difference in the harmonic level between the current statistical period and the previous statistical period, the less relevant the operating parameters of the current statistical period and the previous statistical period are, that is, the lower the correlation degree, and the correlation degree is inversely proportional to the absolute value of the comprehensive difference in the harmonic level. The larger the absolute value of the difference in the unbalance degree of the three-phase current between the current statistical period and the previous statistical period, the less relevant the three-phase currents of the current statistical period and the previous statistical period are, indicating that the correlation degree of the operating parameters between the current statistical period and the previous statistical period is lower, and the correlation degree is inversely proportional to the absolute value of the difference in the unbalance degree of the three-phase current.

[0055] The following gives a specific quantification method for the correlation degree: ; where Indicates the correlation degree of the operating parameters between the (M + 1)-th statistical period and the M-th statistical period. This calculation formula comprehensively analyzes the absolute value of the comprehensive difference in the harmonic level after negative correlation and the absolute value of the difference in the unbalance degree of the three-phase current after negative correlation, and obtains the correlation degree by taking the average value of the absolute value of the comprehensive difference in the harmonic level after negative correlation and the absolute value of the difference in the unbalance degree of the three-phase current. In essence, it is a weighted sum of the two, and the weights are both 0.5.

[0056] Step S4: Determine the reference weight of the harmonic level according to the magnitude relationship between the correlation degree corresponding to the current statistical period and the degree of change in the operation task of the permanent magnet synchronous motor in the current statistical period.

[0057] First, obtain the degree of change in the operation tasks of the permanent magnet synchronous motor in the current statistical period. Specifically, in the current statistical period, each time the operation task of the permanent magnet synchronous motor changes, the change process (i.e., the changed operation task) is recorded, so as to obtain the actual number of types of operation tasks that occur in the permanent magnet synchronous motor in the current statistical period according to the change situation of the operation tasks of the permanent magnet synchronous motor in the current statistical period. The change in the operation task refers to the change in the working mode, processing task, or processing material, etc. of the permanent magnet synchronous motor. And obtain the total number of types of operation tasks of the permanent magnet synchronous motor. The total number of types represents the number of types of all possible operation tasks of the permanent magnet synchronous motor related to the application scenario, which is a known value. Then, the degree of change in the operation tasks of the permanent magnet synchronous motor in the current statistical period is the proportion of the types of operation tasks of the permanent magnet synchronous motor in the current statistical period, and the proportion of types is the ratio of the actual number of types of operation tasks of the permanent magnet synchronous motor in the current statistical period to the total number of types of operation tasks of the permanent magnet synchronous motor.

[0058] Determine the reference weight of the harmonic degree according to the magnitude relationship between the correlation degree corresponding to the current statistical period and the degree of change in the operation tasks of the permanent magnet synchronous motor in the current statistical period. In an exemplary embodiment, it is divided into the following two cases: If the correlation degree between the current statistical period and the previous statistical period is greater than the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period, it means that under the change of the operation task of the permanent magnet synchronous motor in the current statistical period, the operation parameters in the current statistical period have not changed significantly. If there are harmonics in the current statistical period, it is very likely that the harmonic change in the current statistical period is caused by the change of the operation task. For the normal harmonic fluctuation caused by the change of the operation task, the reference weight of the harmonic degree in the current statistical period needs to be set to a smaller value to avoid misjudgment. In an exemplary embodiment, a first preset weight threshold is set. The first preset weight threshold is a relatively small weight threshold, and the numerical range of the first preset weight threshold is (0.2, 0.5]. The specific value of the first preset weight threshold is set according to actual needs. Moreover, the reference weight of the harmonic degree in the current statistical period is less than the first preset weight threshold to limit the reference weight of the harmonic degree in the current statistical period within a small numerical range. Moreover, the gap between the correlation degree between the current statistical period and the previous statistical period and the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period is inversely proportional to the reference weight of the harmonic degree. Specifically: the greater the gap between the correlation degree between the current statistical period and the previous statistical period and the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period, that is, the greater the correlation degree between the current statistical period and the previous statistical period is than the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period, the smaller the reference weight of the harmonic degree. In an exemplary embodiment, a specific quantification method is given, and a parameter is calculated : ; wherein, represents the change degree of the operation task of the permanent magnet synchronous motor in the (M + 1)-th statistical period.

[0059] If the correlation degree between the current statistical period and the previous statistical period is greater than the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period, then is less than 1. As a specific example, when is equal to 1, the reference weight of the harmonic degree in the current statistical period is equal to 0.5 (that is, the first preset weight threshold is set to 0.5); when is less than 1, the reference weight of the harmonic degree in the current statistical period is equal to 0.5 , so that is smaller, the reference weight of the harmonic degree in the current statistical period is smaller. When is less than 0.2, the reference weight of the harmonic degree in the current statistical period is set to 0.2.

[0060] If the correlation degree between the current statistical period and the previous statistical period is less than the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period, it indicates that the change of the operation parameters is too large under the change of the operation task of the permanent magnet synchronous motor in the current statistical period. That is, in addition to the change of the operation task, other factors may also affect the harmonic change in the current statistical period. Therefore, the reference weight of the harmonic degree in the current statistical period needs to be set larger to avoid misjudgment.

[0061] In an exemplary embodiment, a second preset weight threshold is set. The second preset weight threshold is a relatively large weight threshold, and the numerical range of the second preset weight threshold is [0.5, 0.7), and the specific value of the second preset weight threshold is set according to actual needs. Then, the first preset weight threshold is less than or equal to the second preset weight threshold.

[0062] Then the reference weight of the harmonic degree in the current statistical period is greater than the second preset weight threshold, and the gap between the correlation degree corresponding to the current statistical period and the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period is proportional to the reference weight of the harmonic degree. That is, the greater the gap between the correlation degree between the current statistical period and the previous statistical period and the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period, the greater the reference weight of the harmonic degree.

[0063] In an exemplary embodiment, if the correlation degree between the current statistical period and the previous statistical period is less than the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period, then is greater than 1. As a specific example, when is equal to 1, the reference weight of the harmonic degree in the current statistical period is equal to 0.5 (that is, the second preset weight threshold is set to 0.5); when is greater than 1, the reference weight of the harmonic degree in the current statistical period is equal to 0.5 , so that the greater, the greater the reference weight of the harmonic degree in the current statistical period. When it is greater than 0.7, the reference weight of the harmonic degree in the current statistical period is set to 0.7. By adopting the above process, the reference weight of the harmonic degree in the current statistical period is obtained. Among them, the harmonic degree refers to the harmonic degree corresponding to each operation parameter.

[0064]

[0065] ​It should be understood that when training a permanent magnet synchronous motor fault monitoring model, in addition to the harmonic degree, other relevant features of the permanent magnet synchronous motor are obtained, and the specific composition of other features is set according to actual training needs. In an exemplary embodiment, other features of the permanent magnet synchronous motor include: the current mean and current variance of each phase current in the three-phase current during the statistical period, the amplitude, phase and frequency of each phase voltage in the three-phase voltage, the rotational speed average, range and variance, the number of each harmonic of the three-phase current, three-phase voltage and rotational speed, as well as the corresponding amplitude and frequency, and the temperature mean.

[0066] After obtaining the reference weight of the harmonic degree, the reference weight of other features of the permanent magnet synchronous motor is equal to the difference between the value 1 and the reference weight of the harmonic degree. For example, when the reference weight of the harmonic degree is equal to 0.2, the reference weight of other features of the permanent magnet synchronous motor is 0.8; when the reference weight of the harmonic degree is equal to 0.7, the reference weight of other features of the permanent magnet synchronous motor is 0.3. Thus, the reference weights of each feature of the permanent magnet synchronous motor are obtained.

[0067] Step S5: Train the permanent magnet synchronous motor fault monitoring model based on the harmonic degree under the reference weight addition and other features of the permanent magnet synchronous motor.

[0068] Each feature of the permanent magnet synchronous motor corresponds to an input channel, and each input channel has its own reference weight. Each feature is fused with its corresponding reference weight (specifically, each feature is multiplied by its corresponding reference weight) to obtain each feature under the reference weight addition, including: the harmonic degree under the reference weight addition, and other features of the permanent magnet synchronous motor under the reference weight addition.

[0069] Then, in each statistical period, the reference weights of each feature of the permanent magnet synchronous motor are obtained, and the harmonic degree under the reference weight addition corresponding to each statistical period, as well as other features of the permanent magnet synchronous motor under the reference weight addition, can be obtained. Therefore, the features of the permanent magnet synchronous motor under the corresponding reference weight addition in multiple statistical periods are obtained and used to form a training set, which is input into the permanent magnet synchronous motor fault monitoring model for training to obtain the final operation fault detection model of the permanent magnet synchronous motor. The loss function during the training process is the cross-entropy loss function.

[0070] When training the permanent magnet synchronous motor fault monitoring model, manually label whether the permanent magnet synchronous motor corresponding to each sample in the training set has a fault. If there is no operation fault, the output is 0. If there is an operation fault, the output is 1, and the corresponding operation fault name is also labeled. And 20% of the training set is used as the test set.

[0071] When subsequently monitoring the faults of a permanent magnet synchronous motor according to the trained permanent magnet synchronous motor operation fault detection model, the actual data of the permanent magnet synchronous motor collected is input into the model to output the operation state of the permanent magnet synchronous motor. If an operation fault occurs, the name of the operation fault will be output simultaneously. Then, the maintenance personnel will stop the permanent magnet synchronous motor in time, conduct a risk investigation after clarifying the operation fault, and carry out targeted fault diagnosis and treatment.

[0072] This embodiment also provides an operation fault monitoring system applicable to a permanent magnet synchronous motor, including: a memory and a processor; the memory is connected to the processor, and the memory is used to store program instructions; the processor is used to implement the steps in the above-mentioned operation fault monitoring method embodiment applicable to the permanent magnet synchronous motor when the program instructions are executed.

[0073] In an exemplary embodiment, the present invention provides an operation fault monitoring device applicable to a permanent magnet synchronous motor, including: a computer-readable storage medium, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps in the above-mentioned operation fault monitoring method embodiment applicable to the permanent magnet synchronous motor.

[0074] It should be noted that: the above-mentioned sequence of embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0075] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the key points of each embodiment are the differences from other embodiments.

Claims

1. A method for monitoring operation faults applicable to a permanent magnet synchronous motor, characterized in that, Including: Obtain the operating parameters of the permanent magnet synchronous motor in the current statistical period and the previous statistical period, where the operating parameters at least include three-phase current and three-phase voltage; Obtain the harmonic degree of each operating parameter from the number of different harmonics and the abnormal conditions of each harmonic in each operating parameter; According to the change in the harmonic degree between the current statistical period and the previous statistical period, and the change in the unbalance degree of the three-phase current, obtain the correlation degree of the operating parameters between the current statistical period and the previous statistical period; Determine the reference weight of the harmonic degree according to the magnitude relationship between the correlation degree corresponding to the current statistical period and the change degree of the operation task of the permanent magnet synchronous motor in the current statistical period; Train the fault monitoring model of the permanent magnet synchronous motor based on the harmonic degree under the addition of the reference weight and other characteristics of the permanent magnet synchronous motor.

2. The operation fault monitoring method applicable to a permanent magnet synchronous motor according to claim 1, characterized in that, The process of obtaining the harmonic degree of the operating parameter includes: Obtain the fundamental wave and the data groups of each harmonic of the operating parameter, where the data group is composed of amplitude and frequency; Respectively obtain the Euclidean distance between the fundamental wave and the data groups of each harmonic to obtain the comprehensive Euclidean distance; Fuse the comprehensive Euclidean distance and the number of different harmonics in the operating parameter to obtain the harmonic degree of the operating parameter, and the harmonic degree is proportional to both the comprehensive Euclidean distance and the number of different harmonics in the operating parameter.

3. The operation fault monitoring method applicable to a permanent magnet synchronous motor according to claim 1, characterized in that, The process of obtaining the unbalance degree of the three-phase current includes: Obtain the difference between any two-phase currents, where the difference includes the DTW distance and the mean difference of the fluctuation curves of any two-phase currents; Fuse the differences of all any two-phase currents to obtain the unbalance degree of the three-phase current.

4. The operating fault monitoring method applicable to a permanent magnet synchronous motor according to claim 1, characterized in that The change in the harmonic degree between the current statistical period and the previous statistical period is the absolute value of the comprehensive difference of the harmonic degrees between the current statistical period and the previous statistical period, and the absolute value of the comprehensive difference is obtained by fusing the absolute values of the differences in the harmonic degrees of all operating parameters between the current statistical period and the previous statistical period; The change in the unbalance degree of the three-phase current is the absolute value of the difference in the unbalance degree of the three-phase current between the current statistical period and the previous statistical period.

5. The operation fault monitoring method applicable to a permanent magnet synchronous motor according to claim 4, characterized in that, The process of obtaining the correlation degree includes: According to the absolute value of the comprehensive difference in the harmonic degree between the current statistical period and the previous statistical period, and the absolute value of the difference in the unbalance degree of the three-phase current, obtain the correlation degree of the operating parameters between the current statistical period and the previous statistical period; the correlation degree is inversely proportional to both the absolute value of the comprehensive difference in the harmonic degree and the absolute value of the difference in the unbalance degree of the three-phase current.

6. The operating fault monitoring method for a permanent magnet synchronous motor according to claim 1, characterized in that, The change degree of the operation task of the permanent magnet synchronous motor in the current statistical period is the type proportion of the operation task of the permanent magnet synchronous motor in the current statistical period, and the type proportion is the ratio of the actual number of types of the operation task of the permanent magnet synchronous motor in the current statistical period to the total number of types of the operation task of the permanent magnet synchronous motor.

7. The operation fault monitoring method applicable to a permanent magnet synchronous motor according to claim 1, characterized in that The process of obtaining the reference weight of the harmonic degree includes: If the correlation degree corresponding to the current statistical period is greater than the change degree of the operation tasks of the permanent magnet synchronous motor in the current statistical period, the reference weight of the harmonic degree is less than the first preset weight threshold, and the gap between the correlation degree corresponding to the current statistical period and the change degree of the operation tasks of the permanent magnet synchronous motor in the current statistical period is inversely proportional to the reference weight of the harmonic degree; If the correlation degree corresponding to the current statistical period is less than the change degree of the operation tasks of the permanent magnet synchronous motor in the current statistical period, the reference weight of the harmonic degree is greater than the second preset weight threshold, and the gap between the correlation degree corresponding to the current statistical period and the change degree of the operation tasks of the permanent magnet synchronous motor in the current statistical period is directly proportional to the reference weight of the harmonic degree; the first preset weight threshold is less than or equal to the second preset weight threshold.

8. The operation fault monitoring method for a permanent magnet synchronous motor according to claim 1, characterized in that The reference weight of other characteristics of the permanent magnet synchronous motor is equal to the difference between the value 1 and the reference weight of the harmonic degree; Training the permanent magnet synchronous motor fault monitoring model based on the harmonic degree under the reference weight addition and other characteristics of the permanent magnet synchronous motor includes: Obtaining the respective characteristics of the permanent magnet synchronous motor under the corresponding reference weight addition in multiple statistical periods, and training the permanent magnet synchronous motor fault monitoring model therewith.

9. An operating fault monitoring system applicable to a permanent magnet synchronous motor, characterized by comprising: A memory and a processor; The memory is connected to the processor; The memory is used for storing program instructions; The processor is used for implementing the operation fault monitoring method applicable to the permanent magnet synchronous motor according to any one of claims 1-8 when the program instructions are executed.

10. An operating fault monitoring device applicable to a permanent magnet synchronous motor, characterized in that, Including a computer-readable storage medium storing a computer program, which when executed by a processor, implements the steps in the embodiment of the operation fault monitoring method applicable to the permanent magnet synchronous motor according to any one of claims 1-8.

Images