Operation fault monitoring method and monitoring system for vehicle permanent magnet synchronous motor

Abstract

The invention relates to the technical field of permanent magnet synchronous motor fault detection, in particular to an operation fault monitoring method and system for a vehicle permanent magnet synchronous motor, and the method comprises the steps: installing a sensor for the permanent magnet synchronous motor, and collecting the temperature, current and voltage of the permanent magnet synchronous motor in the operation process; the frequency variation degree of the current is calculated through the current operation state of the permanent magnet synchronous motor; obtaining a current extreme value regression degree through change characteristics between extreme points during current operation, and obtaining a current persistence abnormity degree of the permanent magnet synchronous motor; obtaining the voltage persistence abnormity degree and the temperature fault recognition degree of the permanent magnet synchronous motor; and identifying the operation state of the permanent magnet synchronous motor through the neural network. The invention aims to effectively avoid misjudgment of transient abnormal conditions of the permanent magnet synchronous motor caused by environmental interference.

Description

Technical Field

[0001] This application relates to the technical field of permanent magnet synchronous motor fault detection, and specifically relates to a method and system for monitoring the operation faults of a vehicle-mounted permanent magnet synchronous motor. Background Art

[0002] A vehicle-mounted permanent magnet synchronous motor (PMSM) is a synchronous motor with permanent magnets embedded in the rotor. It uses the magnetic field generated by the permanent magnets to interact with the induced magnetic field in the stator winding to generate torque, thereby driving the vehicle. If a permanent magnet synchronous motor fails, such as a rotor position sensor failure or a winding short circuit, it may cause abnormal torque output of the motor. When the vehicle is traveling at high speed, this abnormal torque may cause the vehicle to lose control and trigger serious traffic accidents. Through real-time fault detection, these potential dangerous situations can be detected in time, and corresponding measures can be taken, such as cutting off the motor power supply and activating the braking system, to ensure the safety of the vehicle and passengers.

[0003] For the identification of the operation faults of a permanent magnet synchronous motor, the operation state data is trained through a neural network, and then the operation fault categories of the permanent magnet synchronous motor are identified through the neural network. However, since the vehicle will be in different environments during driving, such as hills, forests, and speed bumps, the vehicle will jolt in such environments, causing changes in the operation state of the permanent magnet synchronous motor and resulting in short-term abnormalities in the collected data. However, the vehicle-mounted permanent magnet synchronous motor has no problems. If the real-time collected data is directly used, the neural network will identify this situation as a fault, resulting in inaccurate identification of the operation faults of the permanent magnet synchronous motor. Summary of the Invention

[0004] In view of the above, it is necessary to provide a method and system for monitoring the operation faults of a vehicle-mounted permanent magnet synchronous motor to solve the above problems.

[0005] The first aspect of this application provides a method for monitoring the operation faults of a vehicle-mounted permanent magnet synchronous motor, and the method includes:

[0006] During a preset time period, collect the temperature during the operation of the permanent magnet synchronous motor to form an operation temperature sequence; at the same time, collect the current and voltage of each phase to form an operation current sequence and an operation voltage sequence of each phase respectively;

[0007] Decompose the operating current sequence of each phase to obtain a preset number of layers of decomposed current sequences. Compare the similarity degree of the frequencies of the current decomposition sequences of each layer after the current decomposition of any two phases, and combine the distribution characteristics of the amplitude values to obtain the relative frequency difference degree of the current decomposition sequences of each layer after the current decomposition of the any two phases. Analyze the numerical characteristics of the relative frequency difference degrees of the current of each phase and the currents of the remaining phases in all layers to obtain the frequency mutation degree of the operating current sequence of each phase;

[0008] Extract and sort the extreme points in the operating current sequence of each phase. Based on the change characteristics between the extreme points, classify the extreme points, and determine the standard characteristic difference of the current of each phase according to the element mean values of all classes; According to the distribution change of the extreme values in the operating current sequence of each phase, combine the standard characteristic difference to obtain the current extreme value regression degree of each phase;

[0009] According to the frequency mutation degree of the currents of all phases and the current extreme value regression degree, obtain the current persistent abnormality degree of the permanent magnet synchronous motor; Correspondingly, analyze the operating voltage sequence of each phase to obtain the voltage persistent abnormality degree of the permanent magnet synchronous motor; Obtain the prediction result of the operating temperature sequence in time series, and combine the current persistent abnormality degree and the voltage persistent abnormality degree to form the operating fault identification vector of the permanent magnet synchronous motor;

[0010] Based on the operating fault identification vector of the permanent magnet synchronous motor, use the trained neural network to obtain the fault monitoring result of the permanent magnet synchronous motor.

[0011] Among them, the obtaining process of the relative frequency difference degree of the current decomposition sequences of each layer after the current decomposition of the any two phases is specifically as follows:

[0012] Calculate the ratio of the number of intersection elements to the number of union elements of the frequencies of the current decomposition sequences of each layer after the current decomposition of any two phases; Calculate the difference between 1 and the ratio, and then multiply it by the mean value of the amplitude values of all non-intersection frequencies to obtain the relative frequency difference degree of the current decomposition sequences of each layer after the current decomposition of the any two phases.

[0013] Among them, the obtaining of the frequency mutation degree of the operating current sequence of each phase is specifically as follows:

[0014] Obtain the minimum value of the relative frequency difference degree of the current decomposition sequences of each layer between the current of each phase and the currents of all other phases, and use the mean value of the minimum values of the relative frequency difference degrees corresponding to each phase in all layers as the frequency mutation degree of the operating current sequence of each phase.

[0015] Among them, the extraction and sorting of the extreme points in the operating current sequence of each phase is specifically as follows: Sort in ascending order based on the numerical values of the extreme points.

[0016] Among them, the specific difference in the standard features of each phase current is specifically the minimum value of the element means of all classes.

[0017] Among them, obtaining the current extreme value regression degree of each phase is specifically as follows:

[0018] For the operating current sequence of each phase, calculate the difference between two adjacent maximum values, denoted as the first difference; denote the difference between the first difference and the standard feature difference as the second difference; take the mean of the second differences corresponding to all adjacent maximum values in the operating current sequence of each phase as the first eigenvalue of each phase.

[0019] Correspondingly, calculate the second eigenvalue of each phase based on adjacent minimum values.

[0020] Based on the first eigenvalue and the second eigenvalue of each phase, obtain the current extreme value regression degree of each phase current. Among them, the current extreme value regression degree is negatively correlated with both the first eigenvalue and the second eigenvalue.

[0021] Among them, the current extreme value regression degree is specifically as follows: Denote the first eigenvalue of the i-th phase as and denote the second eigenvalue of the i-th phase as , then the formula form of the current extreme value regression degree of the i-th phase is: ; where represents the exponential function with the natural constant as the base.

[0022] Among them, obtaining the current persistent abnormality degree of the permanent magnet synchronous motor is specifically as follows:

[0023] For each phase current, calculate the difference between 1 and the current extreme value regression degree, and perform positive fusion with the degree of frequency variation; take the maximum positive fusion result of all phase currents as the current persistent abnormality degree of the permanent magnet synchronous motor.

[0024] Among them, the specific process of training the neural network is as follows:

[0025] Assign different marking values to each fault state of the permanent magnet synchronous motor; obtain the operation fault recognition vectors of the permanent magnet synchronous motor with known faults and the operation recognition vectors of the permanent magnet synchronous motor without faults as the input of the neural network, and the output of the neural network is the marking value corresponding to the fault.

[0026] In a second aspect, an embodiment of the present application further provides an operation fault monitoring system for a vehicle-mounted permanent magnet synchronous motor, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the method described in any one of the above.

[0027] This application has at least the following beneficial effects:

[0028] By collecting the temperature, current, and voltage data during the operation of the permanent magnet synchronous motor, decomposing the current and voltage of each phase, and obtaining the degree of frequency variation according to the frequency distribution characteristics, this application can identify abnormal fluctuations and potential faults during the motor operation, improving the accuracy of fault diagnosis. Further, by extracting the extreme points of the current and voltage during the operation and analyzing the fluctuation conditions of the same type of extreme points, the current extreme value regression degree is obtained, which helps to accurately evaluate the motor operation state and improve the response speed and accuracy of the fault warning system. Further, the current persistent abnormality degree and the voltage persistent abnormality degree are obtained, which can cope with data fluctuations in different environments, enabling the monitoring system to still operate stably under complex working conditions. This stability improves the identification of the persistent abnormal state of the permanent magnet synchronous motor and enhances the practicability and reliability of the monitoring system. By predicting the temperature during the operation and obtaining the temperature fault recognition degree, it helps to identify the overheating problem of the motor or equipment in advance. This warning ability can shorten the fault response time. Compared with traditional methods, it can effectively avoid misjudgment of short-term abnormal conditions caused by environmental interference, accurately identify faults, significantly improve the diagnosis accuracy, and provide reliable guarantee for the safe operation of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flowchart of the steps of a method for monitoring the operation fault of a vehicle-mounted permanent magnet synchronous motor provided by an embodiment of this application;

[0030] Figure 2 It is a schematic diagram for obtaining the current persistent abnormality degree of the permanent magnet synchronous motor provided by an embodiment of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0031] In the description of the embodiments of this application, words such as "exemplary", "or", "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly speaking, using words such as "exemplary", "or", "for example" aims to present relevant concepts in a specific way.

[0032] 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 this application belongs. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0033] In addition, it should be noted that the terms "first" and "second" in this application and the accompanying drawings are used to distinguish similar objects, rather than to describe a specific order or sequence. In the embodiments of this application, the methods disclosed in the method or flowchart include one or more steps for implementing the method. Without departing from the scope of protection of this application, the execution order of multiple steps can be interchanged with each other, and some steps can also be deleted.

[0034] 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 this application belongs.

[0035] The following specifically describes the specific solutions of a method and a monitoring system for monitoring the operating faults of a vehicle-mounted permanent magnet synchronous motor provided by this application with reference to the accompanying drawings.

[0036] Please refer to Figure 1 , which shows a flowchart of the steps of a method for monitoring the operating faults of a vehicle-mounted permanent magnet synchronous motor provided by an embodiment of this application. The method includes the following steps:

[0037] The first step: During a preset time period, collect the temperature during the operation of the permanent magnet synchronous motor to form an operating temperature sequence; at the same time, collect the current and voltage of each phase to form an operating current sequence and an operating voltage sequence for each phase respectively.

[0038] First, install current sensors, voltage sensors, and temperature sensors on the vehicle-mounted permanent magnet synchronous motor. The collection frequency is 1 kHz, and the collection time length is 10 s. It should be noted that the operation of the permanent magnet synchronous motor depends on a three-phase power supply. Therefore, the current and voltage sensors are respectively used to collect the current and voltage data of each phase.

[0039] Since the data collected may be missing during the data collection process, in this application, the mean interpolation method is used to interpolate the missing data. Arrange the interpolated current, voltage, and temperature data in the order of the collection time. Then, use the mean filter algorithm to denoise the arranged data to obtain the operating current sequence, operating voltage sequence, and operating temperature sequence of the permanent magnet synchronous motor. Among them, the calculations of the mean interpolation method and the mean filter are well-known technologies, and the specific calculation steps are not elaborated here.

[0040] Second step: Decompose the operating current sequence of each phase to obtain a preset number of layers of decomposed current sequences. Compare the similarity degree of the frequencies of the current decomposition sequences of each layer after the current decomposition of any two phases, and combine the distribution characteristics of the amplitude values to obtain the relative frequency difference degree of the current decomposition sequences of each layer after the current decomposition of the any two phases. Analyze the numerical characteristics of the relative frequency difference degree of the current of each phase and the currents of the remaining phases in all layers to obtain the frequency mutation degree of the operating current sequence of each phase.

[0041] During the vehicle driving process, due to the uneven road surface, the vehicle will jolt during driving, resulting in vehicle vibration. The vibration will cause various parameters of the permanent magnet synchronous motor of the monitored vehicle to change. Taking the current data as an example, the vehicle vibration may cause the current of the permanent magnet synchronous motor to be in an unstable state. The unstable current will cause the torque fluctuation and speed fluctuation to increase the mechanical stress of the motor, especially the impact force on the bearing, which will exacerbate the bearing wear and shorten the service life of the bearing.

[0042] The input of the permanent magnet synchronous motor is three-phase alternating current obtained through direct and orthogonal transformation. Under ideal conditions, the current waveforms of the three-phase alternating current should be the same, and the phase angles between adjacent two-phase currents differ by 120 degrees. Therefore, the difference of the three-phase currents can reflect the stability of the current. Thus, the operating current sequences of the three phases of the permanent magnet synchronous motor are respectively used as the input of the wavelet decomposition algorithm. For the wavelet basis function of the wavelet decomposition algorithm, in this embodiment, the Symlets8 function is selected, and the decomposition layer number takes the value of 8. The decomposed current data is denoted as the current decomposition sequence. The calculation of the wavelet decomposition algorithm is a well-known technology, and the specific calculation steps are not described in detail here.

[0043] For the current decomposition sequences of the three-phase currents, since the same method is used for decomposition during decomposition, therefore, for the current decomposition sequences of the same layer of the three-phase electricity, under ideal conditions, the distribution of the current data should be the same. Thus, the current decomposition sequences of the same layer of the three-phase currents are used as the input of the Fourier transform algorithm. After transformation, the output result is the frequency of the current decomposition sequence and its corresponding amplitude. It should be noted that the amplitude value is non-zero. The calculation process of the Fourier transform is a well-known technology, and the specific calculation steps are not described in detail here. Since the same signal has the same frequency and corresponding amplitude after Fourier transform. Therefore, when the frequency difference between two current decomposition sequences increases, it means that there are large differences in the currents of different phases of the permanent magnet synchronous motor for a long time, indicating that the possibility of current instability will also increase accordingly.

[0044] Based on this, to calculate the degree of frequency variation of the operating current sequence, first compare the similarity of the frequencies of the current decomposition sequences of each layer after decomposing the currents of any two phases, and combine the overall distribution of the amplitude values of the non-intersecting frequencies to obtain the relative frequency difference of the current decomposition sequences of each layer after decomposing the currents of any two phases. Specifically: calculate the ratio of the number of intersection elements to the number of union elements of the frequencies of the current decomposition sequences of each layer after decomposing the currents of any two phases; calculate the difference between 1 and the ratio, and then multiply it by the mean value of the amplitude values of the non-intersecting frequencies to obtain the relative frequency difference of the current decomposition sequences of each layer after decomposing the currents of any two phases.

[0045] Furthermore, obtain the minimum value of the relative frequency difference of the current decomposition sequences of each layer between the current of each phase and the currents of all other phases, and take the mean value of the minimum values corresponding to all layers of each phase current as the degree of frequency variation of each phase operating current sequence.

[0046] In an ideal situation, the operating current sequences of any two phases of a permanent magnet synchronous motor have the characteristic of the same waveform, and the data of the same layer after wavelet decomposition are the same. Therefore, the frequencies and the corresponding amplitudes obtained through Fourier transform are also the same. When the frequencies of the two decomposed layer data, that is, the current decomposition sequences, differ more, it indicates that there are different stable currents between these two current phases, and the larger the amplitude values of the different frequencies of the two current decomposition sequences, the higher the long-term nature of the difference, the greater the relative fluctuation between the currents, and the higher the instability of the current phase. While the difference between normal current phases is smaller. Therefore, by using the minimum value, the degree of frequency variation of normal current phases can be assigned a smaller value, making the degree of frequency variation of abnormal current phases larger and making it easier to identify abnormal current phases.

[0047] The third step: Extract and sort the extreme points in each phase operating current sequence, classify the extreme points based on the change characteristics between the extreme points, and determine the standard characteristic difference of each phase current according to the mean value of the elements of all classes; according to the distribution change of the extreme values in the operating current sequence of each phase, combine the standard characteristic difference to obtain the current extreme value regression degree of each phase.

[0048] To ensure the stable operation of a vehicle-mounted permanent magnet synchronous motor, the current waveform provided by the direct current converter is usually an ideal sine wave. In this ideal state, the differences between the sine wave peaks of the waveform remain consistent, and the differences between the trough values are also the same. The rotational speed of the motor will be dynamically adjusted according to the driving requirements of the vehicle to achieve precise control and efficient power output. Therefore, the frequency of the three-phase alternating current will also change dynamically accordingly, but the stability of the peak and trough values of the current needs to be maintained. Thus, the operating current sequence of each phase is used as the input of the extreme point detection algorithm, and the output is the extreme points of the operating current sequence of each phase. All the extreme points of each phase current are arranged in ascending order, denoted as the extreme value sequence. Then, the extreme value sequence is used as the input of the first-order difference algorithm, and the output is the extreme value difference sequence of the extreme value sequence. Through the extreme value difference sequence, the regression degree of the current to the peak and trough values can be judged, improving the recognition of the fault state. Among them, the calculations of the extreme point detection algorithm and the first-order difference method are both well-known technologies, and the specific calculation process will not be elaborated here.

[0049] To maintain the stability of the current, the difference between the peaks should be smaller, and the same goes for the trough values. Therefore, the smaller values in the extreme value difference sequence may be the difference between the peaks or the difference between the trough values. Thus, the extreme value difference sequence is used as the input of the segmentation algorithm, and the output is the segmentation set. The segmentation algorithms include OTSU, K-means, mean segmentation algorithm, and median segmentation algorithm. In this embodiment, the OTSU algorithm is adopted. The calculation process of the OTSU algorithm is a well-known technology, and the specific calculation will not be elaborated here. Select the set with the smallest element mean, and use its element mean as the standard feature difference.

[0050] When the peak or trough value belongs to a state with a large difference for a long time, it indicates a higher possibility of a fault in the permanent magnet synchronous motor. Based on this, according to the distribution of the extreme values in the operating current sequence of each phase, combined with the standard feature difference, calculate the current extreme value regression degree of each phase. Specifically: for the operating current sequence of each phase, calculate the difference between two adjacent maximum values, denoted as the first difference; denote the difference between the first difference and the standard feature difference as the second difference; take the mean of the second differences corresponding to all adjacent two maximum values in the operating current sequence of each phase as the first eigenvalue of each phase. Correspondingly, calculate the difference between two adjacent minimum values, denoted as the third difference; denote the difference between the third difference and the standard feature difference as the fourth difference; take the mean of the fourth differences corresponding to all adjacent two minimum values in the operating current sequence of each phase as the second eigenvalue of each phase; based on the first eigenvalue and the second eigenvalue of each phase, obtain the current extreme value regression degree of each phase current.

[0051] In this embodiment, the difference between variables is calculated using the absolute value of the difference; denote the first eigenvalue of the i-th phase as Let the second eigenvalue of the $i$-th phase be denoted as Then the formula form of the current extreme value regression degree of the $i$-th phase is: ; where represents the exponential function with the natural constant as the base.

[0052] It should be understood that the electromagnetic torque of the permanent magnet synchronous motor is closely related to the amplitude of the current. The instability of the current extreme value will cause fluctuations in the electromagnetic torque. When the current peak suddenly increases, the torque will increase instantaneously; when the current peak decreases, the torque will decrease. Torque fluctuations will make the output power of the motor unstable and affect the running smoothness of the motor. By analyzing the differences between adjacent maximum values or adjacent minimum values in the current waveform, as well as the deviations between these differences and the standard feature differences, the variation characteristics of the current can be evaluated. When the difference value increases, it indicates that the possibility of problems occurring during the test is also higher. This means that the fluctuations between adjacent peaks or valleys of the current are more significant, indicating that the current amplitude change of the permanent magnet synchronous motor is larger, causing the first eigenvalue and the second eigenvalue of the current phase to increase, and the change state of the current does not conform to the stable operation state, resulting in a decrease in the value of the current extreme value regression degree of the current phase.

[0053] The fourth step: Obtain the current persistent anomaly degree of the permanent magnet synchronous motor according to the frequency variation degree of all phase currents and the current extreme value regression degree; correspondingly, analyze the operating voltage sequence of each phase to obtain the voltage persistent anomaly degree of the permanent magnet synchronous motor; obtain the prediction result of the operating temperature sequence in time series, and combine the current persistent anomaly degree and the voltage persistent anomaly degree to form the operating fault identification vector of the permanent magnet synchronous motor.

[0054] The currents of the three phases of the permanent magnet synchronous motor are converted from the same direct current, making the current waveforms the same. When a fault occurs in the permanent magnet synchronous motor, the frequency variation degree of each phase current may be small and the fault cannot be identified. However, due to the change in the peak value of its waveform, the current extreme value regression degree of the current phase will decrease. Based on this, the current persistent anomaly degree of the permanent magnet synchronous motor is obtained according to the frequency variation degree of all phase currents and the current extreme value regression degree. Specifically: for each phase current, calculate the difference between 1 and the current extreme value regression degree, and perform positive fusion with the frequency variation degree; take the maximum positive fusion result of all phase currents as the current persistent anomaly degree of the permanent magnet synchronous motor. In this embodiment, the calculation method of multiplying is used to perform positive fusion on multiple variables. Among them, the schematic diagram of obtaining the current persistent anomaly degree of the permanent magnet synchronous motor is as Figure 2 shown.

[0055] It should be understood that the larger the value of the current persistence abnormality of the permanent magnet synchronous motor, the longer the time of the abnormal state of the permanent magnet synchronous motor, and the more likely there are faults such as winding short circuit, insulation aging, and magnet demagnetization inside. At the same time, since the abnormal state changes of different fault currents are different, the different fault types of the permanent magnet synchronous motor can be identified by using the current persistence abnormality.

[0056] Furthermore, the voltage persistence abnormality of the permanent magnet synchronous motor is calculated by using the operating voltage sequence of each phase of the permanent magnet synchronous motor. Then, the operating temperature sequence of the permanent magnet synchronous motor is used as the input of the Exponential Moving Average (EMA) algorithm. The value range of the smoothing parameter of the necessary parameter is [0, 1], and the value in this embodiment is 0.9. The predicted temperature at the next moment output by the EMA algorithm is denoted as the temperature fault recognition degree. The current persistence abnormality, the voltage persistence abnormality, and the temperature fault recognition degree are combined to form an operating fault recognition vector of the permanent magnet synchronous motor. Among them, the calculation of the EMA algorithm is a well-known technology, and the specific calculation steps are not elaborated here.

[0057] The fifth step: Based on the operating fault recognition vector of the permanent magnet synchronous motor, use the trained neural network to obtain the fault monitoring result of the permanent magnet synchronous motor.

[0058] Through the above steps, the operating fault recognition vectors of the vehicle-mounted permanent magnet synchronous motor when a fault occurs are collected. In this embodiment, for each fault type, 1000 operating recognition vectors in the fault state are collected, and 1000 operating recognition vectors in the normal state are also collected. Then, they are marked respectively for each fault state and normal state. In this embodiment, the normal state is marked as 0, and each fault state is given different marked values, such as 1, 2, 3, etc. Each fault state corresponds to a unique marked value, which can be specifically set by the specific implementer. The five-layer BP network is trained by using the training set, and the cross-entropy is used as the loss function, and the AdaGrad optimizer is used to optimize the training result. Then, the operating fault recognition vector of the vehicle-mounted permanent magnet synchronous motor during operation is used as the input of the training model, and the output is the state mark of the vehicle-mounted permanent magnet synchronous motor. Through the marked value, the state of the permanent magnet synchronous motor during the operation of the vehicle is identified. Among them, the calculation of the BP neural network algorithm is a well-known technology, and the specific calculation steps are not elaborated here.

[0059] Based on the same inventive concept as the above method, an embodiment of the present application further provides an operating fault monitoring system for a vehicle permanent magnet synchronous motor, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above methods for monitoring the operating faults of a vehicle permanent magnet synchronous motor.

[0060] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, which contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions marked in the block may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the descriptions. Sometimes, there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. Each block in the block diagram and / or flowchart, as well as the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0061] For those skilled in the art, it is obvious that the present application is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the basic characteristics of the present application. Therefore, from any point of view, the above embodiments of the present application should be regarded as exemplary and non-restrictive. Modifying the technical solutions recorded in the foregoing embodiments, or equivalently replacing some of the technical features, does not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for monitoring the operating faults of a vehicle-mounted permanent magnet synchronous motor, characterized in that, The method includes the following steps: During a preset time period, collect the temperature during the operation of the permanent magnet synchronous motor to form an operating temperature sequence; at the same time, collect the current and voltage of each phase to form an operating current sequence and an operating voltage sequence of each phase respectively; Decompose the operating current sequence of each phase to obtain a preset number of layers of decomposed current sequences. Compare the similarity degree of the frequencies of the current decomposition sequences of each layer after decomposition of any two phases, and combine the distribution characteristics of the amplitude values to obtain the relative frequency difference degree of the current decomposition sequences of each layer after decomposition of the any two phases. Analyze the numerical characteristics of the relative frequency difference degrees of the current of each phase and the currents of the other phases in all layers to obtain the frequency mutation degree of the operating current sequence of each phase; Extract and sort the extreme points in the operating current sequence of each phase. Based on the change characteristics between the extreme points, classify the extreme points. Determine the standard characteristic difference of the current of each phase according to the element mean of all classes; according to the distribution change of the extreme values in the operating current sequence of each phase, combine the standard characteristic difference to obtain the current extreme value regression degree of each phase; According to the frequency mutation degree and the current extreme value regression degree of the currents of all phases, obtain the current persistent abnormality degree of the permanent magnet synchronous motor; correspondingly, analyze the operating voltage sequence of each phase to obtain the voltage persistent abnormality degree of the permanent magnet synchronous motor; obtain the prediction result of the operating temperature sequence in time series, and combine the current persistent abnormality degree and the voltage persistent abnormality degree to form an operating fault recognition vector of the permanent magnet synchronous motor; Based on the operating fault recognition vector of the permanent magnet synchronous motor, use the trained neural network to obtain the fault monitoring result of the permanent magnet synchronous motor.

2. The operating fault monitoring method of a vehicle permanent magnet synchronous motor according to claim 1, wherein, The specific process for obtaining the relative frequency difference degree of the current decomposition sequences of each layer after decomposition of any two phases is as follows: Calculate the ratio of the number of intersection elements to the number of union elements of the frequencies of the current decomposition sequences of each layer after decomposition of any two phases; calculate the difference between 1 and the ratio, and then multiply it by the mean value of the amplitude values of all non-intersection frequencies to obtain the relative frequency difference degree of the current decomposition sequences of each layer after decomposition of the any two phases.

3. The operating fault monitoring method of a vehicle-mounted permanent magnet synchronous motor according to claim 1, characterized in that, The specific process for obtaining the frequency mutation degree of the operating current sequence of each phase is as follows: Obtain the minimum value of the relative frequency difference degree of the current decomposition sequences of each layer between the current of each phase and the currents of all other phases. Take the mean value of the minimum values of the relative frequency difference degrees corresponding to each phase in all layers as the frequency mutation degree of the operating current sequence of each phase.

4. The operating fault monitoring method of a vehicle-mounted permanent magnet synchronous motor according to claim 1, characterized in that The specific process for extracting and sorting the extreme points in the operating current sequence of each phase is as follows: Sort based on the numerical values of the extreme points from small to large.

5. The operating fault monitoring method of a vehicle permanent magnet synchronous motor according to claim 1, characterized in that, The standard characteristic difference of the current of each phase is specifically the minimum value of the element means of all classes.

6. The operating fault monitoring method of a vehicle permanent magnet synchronous motor according to claim 1, characterized in that, The specific process for obtaining the current extreme value regression degree of each phase is as follows: For the operating current sequence of each phase, calculate the difference between two adjacent maximum values, denoted as the first difference; denote the difference between the first difference and the standard characteristic difference as the second difference; take the mean value of the second differences corresponding to all adjacent two maximum values in the operating current sequence of each phase as the first characteristic value of each phase; Correspondingly, calculate the second characteristic value of each phase based on adjacent minimum values; Based on the first eigenvalue and the second eigenvalue of each phase, the current extreme value regression degree of each phase current is obtained, wherein the current extreme value regression degree has a negative correlation with both the first eigenvalue and the second eigenvalue.

7. The operating fault monitoring method of a vehicle-mounted permanent magnet synchronous motor according to claim 6, characterized in that, The current extreme value regression degree is specifically: Denote the first eigenvalue of the i-th phase as , and denote the second eigenvalue of the i-th phase as . Then the formula form of the current extreme value regression degree of the i-th phase is: ; where represents the exponential function with the natural constant as the base.

8. The operating fault monitoring method of a vehicle-mounted permanent magnet synchronous motor according to claim 1, characterized in that, The obtaining of the current persistent abnormality degree of the permanent magnet synchronous motor is specifically as follows: For each phase current, calculate the difference between 1 and the current extreme value regression degree, and perform positive fusion with the degree of frequency variation; take the maximum positive fusion result of all phase currents as the current persistent abnormality degree of the permanent magnet synchronous motor.

9. The operating fault monitoring method for a vehicle-mounted permanent magnet synchronous motor according to claim 1, wherein, The specific process of training the neural network is as follows: Different marker values are respectively assigned to each fault state of the permanent magnet synchronous motor; obtain the operation fault identification vector of the permanent magnet synchronous motor with known faults and the operation identification vector of the permanent magnet synchronous motor without faults as the input of the neural network, and the output of the neural network is the marker value corresponding to the fault.

10. An operating fault monitoring system for a vehicle permanent magnet synchronous motor, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1-9 are implemented.

Images