A high-frequency electrical signal feature extraction method and system for an EMB system

Abstract

This application relates to the field of feature extraction technology, specifically to a method and system for high-frequency electrical signal feature extraction in EMB systems. The method includes: real-time acquisition of phase current data of a three-phase permanent magnet synchronous motor (PMSM) in an EMB system during a first time period, and the zero-sequence voltage signal of the PMSM after injecting a high-frequency current signal into the stator windings of the PMSM during a second time period; dividing the first time period into cycles, obtaining the waveform imbalance of each cycle to obtain characteristic quantities for inter-turn short-circuit fault diagnosis; obtaining the peak-valley difference coefficient of each cycle to obtain characteristic quantities for local demagnetization fault diagnosis; and obtaining the consistency factor and comprehensive diagnostic characteristic value for inter-turn short-circuit fault diagnosis and local demagnetization fault diagnosis, respectively. This application aims to ensure the effective extraction of fault-related feature information under various operating conditions by integrating phase current analysis methods with zero-sequence voltage response analysis methods based on high-frequency current signal injection.

Description

Technical Field

[0001] This application relates to the field of feature extraction technology, specifically to a method and system for extracting high-frequency electrical signal features for EMB systems. Background Technology

[0002] Electro-mechanical braking (EMB) systems, as a fully drive-by-wire intelligent braking system, use electrical signals to control a three-phase permanent magnet synchronous motor to generate braking force, replacing hydraulic oil, compressed air, and complex pipelines in traditional braking systems. It is a key technology for achieving advanced autonomous driving and intelligent chassis. The three-phase permanent magnet synchronous motor (PMSM), as the core drive component of the EMB system, directly determines driving safety based on its health status.

[0003] Three-phase permanent magnet synchronous motors generally consist of two parts: stator and rotor. Stator winding inter-turn short circuit faults and rotor local demagnetization faults are common stator and rotor faults, respectively. Both can cause a sudden increase in motor phase current and an imbalance in three-phase current, which in turn can lead to excessively high local temperature inside the motor, a decrease in output torque, and may cause EMB system failure, thereby causing serious safety accidents.

[0004] The existing EMB system's three-phase permanent magnet synchronous motors depend on the motor's operating conditions. Under low speed and light load conditions, the high-frequency electrical signals caused by faults are very weak and easily masked by factors such as sensor noise, measurement errors, and inverter nonlinearity. Furthermore, the fault electrical signal characteristics of stator winding inter-turn short circuit faults and rotor local demagnetization faults are similar, which can easily lead to misdiagnosis of faults in the three-phase permanent magnet synchronous motors of the EMB system. Summary of the Invention

[0005] In view of the above, it is necessary to provide a method and system for high-frequency electrical signal feature extraction in EMB systems. Compared with traditional high-frequency electrical signal feature extraction methods for EMB systems, by fusing phase current analysis and zero-sequence voltage response analysis based on high-frequency current signal injection, fault-related feature information can be effectively extracted under various operating conditions.

[0006] In a first aspect, embodiments of this application provide a method for extracting high-frequency electrical signal features for an EMB system, the method comprising the following steps:

[0007] Real-time acquisition of phase current data of the three-phase permanent magnet synchronous motor of the EMB system during the first preset time period, and zero-sequence voltage signal of the three-phase permanent magnet synchronous motor after injecting high-frequency current signal into the stator winding of the three-phase permanent magnet synchronous motor during the second preset time period.

[0008] The first time period is divided into cycles. By comparing the distribution differences of current data of different phases in each cycle, the waveform imbalance of each cycle is obtained, and then the characteristic quantity for inter-turn short circuit fault diagnosis is obtained.

[0009] The peak-valley difference of current data in each phase within each cycle is measured. By measuring the similarity of the peak-valley difference of different phases within each cycle and the dispersion of the peak-valley difference of each phase within each cycle, the peak-valley difference coefficient of each cycle is obtained, and then the characteristic quantity for local demagnetization fault diagnosis is obtained.

[0010] By analyzing the discreteness of waveform imbalance and peak-valley difference coefficients for all cycles, consistency factors for inter-turn short-circuit fault diagnosis and local demagnetization fault diagnosis are obtained.

[0011] By statistically analyzing the injection frequency of the high-frequency current signal and the offset frequency when a local demagnetization fault occurs under the current operating conditions, combined with the amplitude value in the spectrum of the zero-sequence voltage signal, and the characteristic quantities and consistency factors of inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis, comprehensive diagnostic characteristic values ​​for inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis are obtained respectively.

[0012] In one embodiment, the process of obtaining the waveform imbalance is as follows:

[0013] Obtain the peak and trough points of the current data of each phase in each cycle in the time sequence, and calculate the difference between each peak point and its adjacent trough point; take the average of all the corresponding difference values ​​in each cycle as the average peak-trough distance of each phase in each cycle.

[0014] Arrange the average peak-to-valley distances of all phases within each period in descending order. The expression for the waveform imbalance of each period is:

[0015] In the formula, This represents the waveform imbalance in the i-th cycle; This represents the maximum value among the average peak-to-valley distances of all phases within the i-th period; This represents the median value of the average peak-to-valley distance among all phases within the i-th period; This represents the minimum value among the average peak-to-valley distances of all phases within the i-th period; This indicates a preset positive number.

[0016] In one embodiment, the characteristic quantity of the inter-turn short-circuit fault diagnosis is the mean of the waveform imbalance over all cycles.

[0017] In one embodiment, the process of obtaining the peak-valley difference coefficient is as follows:

[0018] The similarity of the out-of-phase waveforms in each period is obtained by measuring the similarity of the difference values ​​between different phases within each period.

[0019] The volatility of the in-phase waveform in each period is obtained by measuring the degree of dispersion of the corresponding difference values ​​within each period.

[0020] The similarity of the out-of-phase waveform and the volatility of the in-phase waveform are respectively mapped to a first positive number and a second positive number;

[0021] The peak-valley difference coefficient is the product of the first positive number and the second positive number.

[0022] In one embodiment, the process of obtaining the out-of-phase waveform similarity is as follows:

[0023] Calculate the similarity of the difference values ​​between any two phases in each cycle, wherein the similarity of the out-of-phase waveforms is the mean of the similarities between all any two phases in each cycle.

[0024] In one embodiment, the process of obtaining the in-phase waveform ripple is as follows:

[0025] The dispersion of all the difference values ​​of each phase within each period is denoted as the peak-valley distance dispersion of each phase within each period;

[0026] The in-phase waveform volatility is the average of the peak-to-valley distance dispersion of all phases within each period.

[0027] In one embodiment, the characteristic quantity of the local demagnetization fault diagnosis is the mean of the peak-valley difference coefficients for all periods.

[0028] In one embodiment, the method for obtaining the consistency factor for inter-turn short-circuit fault diagnosis and local demagnetization fault diagnosis is as follows:

[0029] The consistency factor for inter-turn short-circuit fault diagnosis is the reciprocal of the sum of the dispersion of waveform imbalance in all cycles and a preset constant greater than 0.

[0030] The consistency factor for the diagnosis of local demagnetization faults is the reciprocal of the sum of the dispersion of the peak-valley difference coefficients of all periods and a preset constant greater than 0.

[0031] In one embodiment, the expressions for the combined diagnostic feature values ​​of the inter-turn short-circuit fault diagnosis and the local demagnetization fault diagnosis are as follows:

[0032] In the formula, This represents the comprehensive diagnostic characteristic value for inter-turn short-circuit fault diagnosis; Normalized values ​​of characteristic quantities representing inter-turn short-circuit fault diagnosis; This represents the normalized value of the consistency factor for inter-turn short-circuit fault diagnosis. The normalized value of the amplitude at the injection frequency in the spectrum of the zero-sequence voltage signal;

[0033] In the formula, The comprehensive diagnostic feature value represents the diagnosis of local demagnetization faults; Normalized values ​​of characteristic quantities representing local demagnetization fault diagnosis; This represents the normalized value of the consistency factor for diagnosing local demagnetization faults. This represents the normalized value of the amplitude at the offset frequency in the spectrum of the zero-sequence voltage signal.

[0034] Secondly, embodiments of this application also provide a high-frequency electrical signal feature extraction system for an EMB system, 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 of the above-described high-frequency electrical signal feature extraction methods for an EMB system.

[0035] This application has at least the following beneficial effects:

[0036] This application considers that inter-turn short-circuit faults can cause an imbalance in the waveform of current data between the faulty and non-faulty phases. By dividing the current data into different periods and comparing the distribution differences of current data in different phases within different periods, this waveform imbalance can be accurately captured, thus obtaining a characteristic quantity that can effectively characterize inter-turn short-circuit faults. Local demagnetization faults can cause the current data of each phase of a three-phase permanent magnet synchronous motor to show certain similarities and discrete changes in peak-valley differences. By obtaining the peak-valley difference coefficient, characteristic information related to local demagnetization faults can be extracted from the variation law of current data, making local demagnetization faults distinguishable from other faults in terms of characteristics, and enhancing the ability to distinguish faults in fault diagnosis.

[0037] Furthermore, by assessing the consistency of inter-turn short-circuit faults and local demagnetization faults within different cycles, the reliability of the extracted feature quantities of inter-turn short-circuit faults and local demagnetization faults is evaluated. Under complex operating conditions such as low speed, light load, and high noise, it can effectively filter false alarms caused by various interference factors, thereby enhancing the robustness of feature extraction and fault diagnosis of high-frequency electrical signals of three-phase permanent magnet synchronous motors in the EMB system.

[0038] Furthermore, by fusing phase current analysis with zero-sequence voltage response analysis based on high-frequency current signal injection, feature extraction of high-frequency electrical signals of early faults in three-phase permanent magnet synchronous motors of EMB systems can be performed. This allows for the fusion of various fault-related information, enabling the effective extraction of fault-related feature information under various operating conditions. This helps to detect potential faults in three-phase permanent magnet synchronous motors of EMB systems in their early stages, providing strong support for timely measures to ensure driving safety. Attached Figure Description

[0039] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 A flowchart illustrating the steps of a high-frequency electrical signal feature extraction method for an EMB system, provided in one embodiment of this application;

[0041] Figure 2 This is a schematic diagram illustrating the process of obtaining the consistency factor.

[0042] Figure 3 A schematic diagram illustrating the process of obtaining comprehensive diagnostic feature values ​​for inter-turn short-circuit fault diagnosis;

[0043] Figure 4 A schematic diagram illustrating the process of obtaining comprehensive diagnostic feature values ​​for diagnosing local demagnetization faults. Detailed Implementation

[0044] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. It should be understood that, unless otherwise stated, " / " in this application means "or".

[0046] It should also be noted that the terms "first" and "second" in this application are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0047] The following description, in conjunction with the accompanying drawings, details a specific scheme for a high-frequency electrical signal feature extraction method and system for an EMB system provided in this application.

[0048] Please see Figure 1 The diagram illustrates a flowchart of a high-frequency electrical signal feature extraction method for an EMB system according to an embodiment of this application. The method includes the following steps:

[0049] Step 1: Real-time acquisition of phase current data of the three-phase permanent magnet synchronous motor of the EMB system during the first preset time period, and zero-sequence voltage signal of the three-phase permanent magnet synchronous motor after injecting high-frequency current signal into the stator winding of the three-phase permanent magnet synchronous motor during the second preset time period.

[0050] The automotive EMB system completely eliminates the hydraulic circuit, directly driving the actuators to generate clamping force via wheel-side motors. The EMB system's actuators comprise five main modules: a service brake mechanism, a parking brake mechanism, a brake clearance compensation mechanism, a rapid retraction mechanism, and sensors. The service brake mechanism, as the basic actuator module, mainly consists of a three-phase permanent magnet synchronous motor, a reduction and force amplification mechanism, a motion conversion mechanism, and a pressing component. The braking torque of the EMB system is highly dependent on the performance of the three-phase permanent magnet synchronous motor, and the health of the motor directly determines driving safety.

[0051] Because the zero-sequence voltage of a faulty motor exhibits more pronounced fault characteristics under high-frequency current signal injection, a high-frequency current signal is injected into the stator winding of the three-phase permanent magnet synchronous motor through a converter connected to the stator winding. A voltage sensor is placed between the neutral point of the resistor network and the neutral point of the stator winding of the three-phase permanent magnet synchronous motor to obtain the zero-sequence voltage of the three-phase permanent magnet synchronous motor in the EMB system. Current sensors are arranged on the three-phase windings of the three-phase permanent magnet synchronous motor in the EMB system to obtain the current data of each phase of the three-phase permanent magnet synchronous motor. Both the current sensor and the voltage sensor are closed-loop Hall sensors. In this embodiment, the sampling frequency of both the voltage sensor and the current sensor is set to 10kHz. The sampling frequency value is preset by the user, and the implementer can set it according to the actual situation; this application does not impose any special restrictions. Table 1 shows the parameters of the three-phase permanent magnet synchronous motor in the EMB system.

[0052] Table 1 Parameters of the three-phase permanent magnet synchronous motor in the EMB system

[0053]

[0054] The high-frequency electrical signal acquisition method for a three-phase permanent magnet synchronous motor is as follows:

[0055] (1) During the 0-t1 period, the three-phase permanent magnet synchronous motor of the EMB system operates under the conditions of speed v and rated current, and the current data of each phase of the three-phase permanent magnet synchronous motor of the EMB system are collected in real time.

[0056] In this embodiment, the length of the 0-t1 time period is 2s. The length of the 0-t1 time period is preset by the user and can be set by the user according to the actual situation. This application does not impose any special restrictions. Since the vehicle may frequently make minor braking adjustments during normal driving, the three-phase permanent magnet synchronous motor of the EMB system is in a typical low-speed state at this time, and the speed v is 300 rpm. The speed v is preset by the user and can be set by the user according to the actual situation. This application does not impose any special restrictions.

[0057] (2) During the t1-t2 period, a rotating high-frequency current signal is injected into the stator winding of the three-phase permanent magnet synchronous motor of the EMB system, and the motor is operated under the same working conditions as the 0-t1 period. The zero-sequence voltage signal of the three-phase permanent magnet synchronous motor of the EMB system is collected in real time.

[0058] In this embodiment, the length of the t1-t2 time period is 2.4s. The length of the t1-t2 time period is preset by the user, and the implementer can set it according to the actual situation. This application does not impose any special restrictions. The injection frequency of the high-frequency current signal... The value is 500Hz, and the amplitude I of the high-frequency current signal is 0.2A. The injection frequency is... The values ​​of amplitude I are preset by humans, and implementers can set them according to the actual situation. This application does not impose any special restrictions.

[0059] Step 2: Obtain the feature quantities for inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis respectively; obtain the consistency factors for inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis respectively.

[0060] Three-phase permanent magnet synchronous motors generally consist of two parts: a stator and a rotor. Stator winding inter-turn short circuit faults and rotor local demagnetization faults are common stator and rotor faults, respectively. Both can cause a sudden increase in phase current and imbalance of three-phase current in the three-phase permanent magnet synchronous motor, which in turn leads to excessively high local temperature inside the three-phase permanent magnet synchronous motor, a decrease in output torque, and may cause EMB system failure, thereby causing serious safety accidents.

[0061] Step 2.1: Divide the first time period into cycles. By comparing the distribution differences of current data of different phases in each cycle, obtain the waveform imbalance of each cycle, and then obtain the characteristic quantity for inter-turn short circuit fault diagnosis.

[0062] For three-phase permanent magnet synchronous motors in EMB systems, inter-turn short circuits are a typical early motor fault caused by insulation degradation. They mainly manifest as a local short circuit between several adjacent turns in the stator winding. If they cannot be effectively identified and intervened in the early stage of inter-turn short circuit faults, they are very likely to evolve into more serious single-phase ground faults or phase-to-phase short circuits, ultimately leading to the overall damage of the three-phase permanent magnet synchronous motor and the loss of EMB braking system function.

[0063] An inter-turn short circuit is equivalent to adding a loop to the faulty phase of the motor winding in the EMB system. If phase A is the faulty phase, the short-circuit turns ratio will affect the current data of phase A, but have a smaller impact on the current data of phases B and C. The current data of phases B and C are highly similar, resulting in an imbalance in the waveform of the current data of the three phases A, B and C.

[0064] Since the acquired current data and zero-sequence voltage signal are both high-frequency electrical signals, the 0-t1 time period is divided into several cycles. Based on the above analysis, the waveform imbalance of each cycle is obtained by comparing the distribution differences of current data in different phases within each cycle. The specific process is as follows:

[0065] Obtain the peak and trough points of the current data of each phase in each cycle in the time sequence, calculate the difference between each peak point and its adjacent trough point, and take the average of all the corresponding difference values ​​in each cycle as the average peak-trough distance of each phase in each cycle.

[0066] Arrange the average peak-to-valley distances of all phases within each period in descending order; the expression for the waveform imbalance of each period is:

[0067] In the formula, This represents the waveform imbalance in the i-th cycle; This represents the maximum value among the average peak-to-valley distances of all phases within the i-th period; This represents the median value of the average peak-to-valley distance among all phases within the i-th period; This represents the minimum value among the average peak-to-valley distances of all phases within the i-th period; This indicates a preset positive number, used to avoid a denominator of 0, and also to avoid... The value of affects the calculation results of waveform imbalance. It should be an extremely small positive number; in this embodiment... The value is 0.01, and the implementer can set it according to the actual situation. The specific value to be taken.

[0068] In this embodiment, the period length is 50ms. The period length is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.

[0069] In this embodiment, the AMPD (Automatic Multiscale-based Peak Detection) algorithm is used to obtain the peak and trough points of the current data in the time series. The AMPD algorithm is a well-known technology and will not be described in detail in this application. As other implementation methods, based on the ability to obtain the peak and trough points of the current data in the time series, implementers may use other existing technologies, such as peak and trough detection algorithms, extreme point detection algorithms, etc. This application does not impose any special restrictions.

[0070] In this embodiment, the difference between each peak point and its subsequent trough point is calculated. If there is no trough point after any peak point in the current data of any phase in any period, the difference between the peak point and the average value of all trough points in the current data of any phase in any period is calculated.

[0071] In this embodiment, the difference between the peak and the trough is the absolute value of the difference.

[0072] It should be noted that: Used to quantify the degree of difference between non-faulty phases, in inter-turn short-circuit faults, only the current amplitude of the faulty phase increases significantly. The less the current amplitude of the non-faulty phases is affected by the inter-turn short-circuit fault, the more prominent the change in the current amplitude of the faulty phase becomes. In inter-turn short-circuit faults, the change in the current amplitude of the faulty phase will increase significantly, resulting in its average peak-to-valley distance being much larger than the current amplitudes of the other two phases. The larger the value, the more comprehensively the waveform imbalance during the operation of the three-phase permanent magnet synchronous motor in the EMB system can be captured.

[0073] If an inter-turn short circuit fault occurs in the rotor of a three-phase permanent magnet synchronous motor in an EMB system, the waveform imbalance of the motor in each cycle during the 0-t1 time period is relatively high. The average value of the waveform imbalance in all cycles during the 0-t1 time period is taken as the characteristic quantity of the inter-turn short circuit fault.

[0074] Step 2.2: Measure the peak-valley difference of the current data of each phase in each cycle. By measuring the similarity of the peak-valley differences of different phases in each cycle and the dispersion of the corresponding peak-valley differences in each cycle, obtain the peak-valley difference coefficient of each cycle, and then obtain the characteristic quantity for local demagnetization fault diagnosis.

[0075] If the three-phase permanent magnet synchronous motor of the EMB system operates in a harsh working environment, it is subject to high temperature rise, demagnetizing magnetic field, physical impact, etc. The permanent magnet material on the rotor of the three-phase permanent magnet synchronous motor is prone to local demagnetization fault, causing the permeability curve to deviate and the magnetic flux density distribution in the air gap of the motor to be distorted. This directly leads to a decrease in the no-load back EMF of the three-phase permanent magnet synchronous motor. At this time, in order to maintain the electromagnetic torque output of the three-phase permanent magnet synchronous motor, it is necessary to increase the stator current, which further aggravates the demagnetization of the permanent magnet. In severe cases, it will permanently damage the three-phase permanent magnet synchronous motor and cause the braking failure of the EMB system.

[0076] When a local demagnetization fault occurs in the rotor of a three-phase permanent magnet synchronous motor in an EMB system, the single-phase back electromotive force will decrease as the degree of demagnetization deepens, resulting in an increase in single-phase current. However, when a local demagnetization fault occurs in the rotor of a three-phase permanent magnet synchronous motor, the impact on the stator phase current is similar, and the local demagnetization fault will cause the current amplitude of each phase to be unequal.

[0077] Based on the above analysis, the similarity of the out-of-phase waveforms in each period is obtained by measuring the similarity of the difference values ​​between different phases within each period. The expression is as follows:

[0078] In the formula, This represents the similarity of the out-of-phase waveforms in the i-th period; all the corresponding difference values ​​within each period are used to form the set of peak-valley distances for each phase within each period. This represents the similarity of the set of peak-valley distances between phase A and phase B within the i-th period; This represents the similarity of the set of peak-valley distances between phase A and phase C within the i-th period; This represents the similarity of the set of peak-valley distances between phase B and phase C within the i-th period.

[0079] In this embodiment, the similarity between the peak-valley distance sets is specifically the Jaccard similarity coefficient. The Jaccard similarity coefficient is a well-known technology and will not be described in detail here. As other implementation methods, based on the ability to measure the similarity between the peak-valley distance sets, the implementer may use other existing technologies, such as the Sørensen-Dice coefficient, etc. This application does not impose any special restrictions.

[0080] It should be noted that: through The similarity of the peak-valley distances of the three-phase currents in a three-phase permanent magnet synchronous motor is measured as a whole. If there is a local demagnetization fault in the three-phase permanent magnet synchronous motor of the EMB system, it will cause distortion of the air gap magnetic field. The more consistent the impact of the local demagnetization fault on the stator phase currents as a whole, the higher the similarity of the three-phase waveforms.

[0081] Furthermore, by analyzing the dispersion of the corresponding difference values ​​within each period, the volatility of the in-phase waveform in each period is obtained, expressed as:

[0082] In the formula, This represents the in-phase waveform variability in the i-th period; This represents the dispersion of all the difference values ​​of phase A within the i-th period; This represents the dispersion of all the difference values ​​of phase B within the i-th period; This represents the dispersion of all the difference values ​​of phase C within the i-th period.

[0083] It should be noted that the dispersion of data refers to the degree of unevenness in the distribution of data, which can be achieved by calculating the standard deviation, variance, coefficient of variation, etc. This application does not impose any special restrictions on this.

[0084] In this embodiment, the dispersion of the difference values ​​is the standard deviation.

[0085] It should be noted that: through To measure the waveform change characteristics of each phase of a three-phase permanent magnet synchronous motor, demagnetization causes distortion of the air gap magnetic flux density distribution, resulting in distortion of the back electromotive force waveform. In order to maintain torque output, the amplitude of the current in each phase will change unequally, increasing the volatility of the single-phase current waveform.

[0086] Furthermore, if the rotor of the three-phase permanent magnet synchronous motor in the EMB system experiences a local demagnetization fault, the similarity of out-of-phase waveforms and the volatility of in-phase waveforms are both high in each cycle within the 0-t1 time period. Therefore, the similarity of out-of-phase waveforms and the volatility of in-phase waveforms in each cycle are mapped to a first positive number and a second positive number, respectively. The product of the first positive number and the second positive number is used as the peak-valley difference coefficient for each cycle. The average of the peak-valley difference coefficients for all cycles within the 0-t1 time period is used as the characteristic quantity of the local demagnetization fault. The purpose of mapping the out-of-phase waveform similarity to a positive number is to avoid the situation where the calculated result of the peak-valley difference coefficient is forced to be 0 when the out-of-phase waveform similarity is 0; the purpose of mapping the in-phase waveform volatility to a positive number is to avoid the situation where the calculated result of the peak-valley difference coefficient is forced to be 0 when the in-phase waveform volatility is 0.

[0087] It should be noted that there are many ways to map data to positive numbers. Specifically, it can be achieved by calculating the sum of the data and a value greater than 0, or by using the data as the exponent of an exponential function with the natural constant as the base. This application does not impose any special restrictions on this.

[0088] In this embodiment, the similarity of the out-of-phase waveforms is calculated and compared with a preset value greater than 0. The summation maps the similarity of out-of-phase waveforms to positive numbers; the similarity is calculated by comparing the ripple of in-phase waveforms with a preset value greater than 0. The sum of these values ​​maps the fluctuation of in-phase waveforms to positive numbers; where a preset value greater than 0 is used. Preset values ​​greater than 0 The value of is 0.01. Implementers can set its specific value according to the actual situation. This application does not impose any special restrictions.

[0089] Step 2.3: Obtain the consistency factors for inter-turn short-circuit fault diagnosis and local demagnetization fault diagnosis by considering the discreteness of waveform imbalance and peak-valley difference coefficient for all cycles.

[0090] The condition monitoring of the three-phase permanent magnet synchronous motor in the EMB system depends on the motor's operating conditions. Under low-speed and light-load conditions, the high-frequency electrical signals caused by faults are very weak and easily masked by factors such as sensor noise, measurement errors, and inverter nonlinearity. The current data of each phase collected contains high-order harmonics and interference, resulting in distorted current data. This is very likely to affect the effectiveness of the peak-valley difference coefficient and waveform imbalance, making it difficult to accurately identify local demagnetization faults and inter-turn short-circuit faults in the three-phase permanent magnet synchronous motor of the EMB system.

[0091] The local demagnetization fault and inter-turn short circuit fault of the three-phase permanent magnet synchronous motor in the EMB system have a limited degree of aggravation in a very short time. Therefore, the dispersion of the waveform imbalance of all cycles within the 0-t1 time period is compared with a preset constant greater than 0. The reciprocal of the sum is used as a consistency factor for inter-turn short-circuit fault diagnosis. Higher consistency in waveform imbalance across multiple cycles within the 0-t1 time period indicates lower distortion in the current data of the three-phase permanent magnet synchronous motor in the EMB system. This allows for better capture of inter-turn short-circuit fault characteristics through phase current analysis, leading to higher reliability in diagnosing inter-turn short-circuit faults in the EMB system using these characteristic quantities. Simultaneously, the dispersion of the peak-valley difference coefficient across all cycles within the 0-t1 time period is compared with a preset constant greater than 0. The reciprocal of the sum is used as a consistency factor for diagnosing local demagnetization faults. A higher consistency in the peak-valley difference coefficients across multiple cycles within the 0-t1 time period indicates lower distortion in the current data of the three-phase permanent magnet synchronous motor in the EMB system. This allows for better capture of local demagnetization fault characteristics through phase current analysis, leading to higher reliability in diagnosing local demagnetization faults in the three-phase permanent magnet synchronous motor of the EMB system using these characteristic quantities. A preset constant greater than 0 is used. Preset constants greater than 0 Both are used to avoid a denominator of 0, and a constant greater than 0 is preset. Preset constants greater than 0 The values ​​are all preset by humans, and the implementer can set them according to the actual situation. In this embodiment, a constant greater than 0 is preset. Preset constants greater than 0 The value of each is 0.01. A schematic diagram of the process for obtaining the consistency factor is shown below. Figure 2 As shown.

[0092] In this embodiment, the dispersion of waveform imbalance is the coefficient of variation, and the dispersion of peak-valley difference coefficient is the coefficient of variation.

[0093] Step 3: Calculate the injection frequency of the high-frequency current signal and the offset frequency when a local demagnetization fault occurs under the current operating conditions. Combine the amplitude value in the spectrum of the zero-sequence voltage signal to obtain the characteristic values ​​for inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis, respectively.

[0094] Injecting a non-destructive high-frequency current signal into the stator windings of a three-phase permanent magnet synchronous motor (PMSM) causes the motor impedance to change with the rotor position due to the salient polarity of the PMSM, i.e., the asymmetry in the magnetic circuit. Simultaneously, the inductance parameters of the PMSM are also affected. When a PMSM experiences inter-turn short-circuit faults or partial demagnetization faults, a zero-sequence voltage response analysis method based on high-frequency current signal injection is used for fault diagnosis. Its diagnostic performance is unaffected by the operating conditions of the PMSM within the EMB system, maintaining stable diagnostic performance even under low speed and low current conditions.

[0095] If an inter-turn short-circuit fault occurs in a three-phase permanent magnet synchronous motor of an EMB system, a fault component caused by the short-circuit current will appear in the zero-sequence voltage, and the fault component of the zero-sequence voltage will be related to the injection frequency of the high-frequency current signal. Same frequency; if a partial demagnetization fault occurs in the three-phase permanent magnet synchronous motor of the EMB system, the fault component of the zero-sequence voltage will have a frequency shift from the injected frequency, and the frequency shift is... ,in, The fundamental frequency of the three-phase permanent magnet synchronous motor in the EMB system is calculated using the following formula: In the formula, P represents the number of poles of the three-phase permanent magnet synchronous motor in the EMB system. This indicates the operating speed of the three-phase permanent magnet synchronous motor in the EMB system. It is denoted as the offset frequency.

[0096] Obtain the spectrum of the zero-sequence voltage signal of the three-phase permanent magnet synchronous motor of the EMB system during the time period t1-t2, and record the amplitude value at the injection frequency in the spectrum. Record the amplitude value at the offset frequency in the spectrum. .

[0097] In this embodiment, the Fast Fourier Transform (FFT) algorithm is used to obtain the spectrum of the zero-sequence voltage signal. The FFT algorithm is a well-known technology and will not be described in detail here. As other implementation methods, based on the ability to obtain the spectrum of the zero-sequence voltage signal, implementers may use other existing feasible technologies. This application does not impose any special restrictions.

[0098] Theoretically, when a three-phase permanent magnet synchronous motor in an EMB system experiences an inter-turn short-circuit fault, the spectrum of the zero-sequence voltage signal does not contain a frequency of [frequency missing]. The amount, therefore, As a characteristic value for inter-turn short-circuit fault diagnosis, As a characteristic value for identifying local demagnetization faults.

[0099] Step 4: Obtain the comprehensive diagnostic feature values ​​for inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis by using the feature values ​​of inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis, as well as the feature quantities and consistency factors of inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis.

[0100] Based on steps 1, 2, and 3, for each of the M EMP system's three-phase permanent magnet synchronous motor samples, the characteristic quantities and consistency factors for inter-turn short-circuit fault diagnosis, the characteristic quantities and consistency factors for local demagnetization fault diagnosis, the characteristic values ​​for inter-turn short-circuit fault diagnosis, and the characteristic values ​​for local demagnetization fault diagnosis are calculated.

[0101] In this embodiment, the value of M is 50. The value of M is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.

[0102] Furthermore, the expressions for the comprehensive diagnostic characteristic values ​​of inter-turn short-circuit fault diagnosis and local demagnetization fault diagnosis are as follows:

[0103] In the formula, This represents the comprehensive diagnostic characteristic value for inter-turn short-circuit fault diagnosis; Normalized values ​​of characteristic quantities representing inter-turn short-circuit fault diagnosis; This represents the normalized value of the consistency factor for inter-turn short-circuit fault diagnosis. This represents the normalized value of the amplitude at the injection frequency in the spectrum of the zero-sequence voltage signal, i.e., the normalized value of the characteristic value for inter-turn short-circuit fault diagnosis; a schematic diagram of the process for obtaining the comprehensive diagnostic characteristic value for inter-turn short-circuit fault diagnosis is shown below. Figure 3 As shown.

[0104] In the formula, The comprehensive diagnostic feature value represents the diagnosis of local demagnetization faults; Normalized values ​​of characteristic quantities representing local demagnetization fault diagnosis; This represents the normalized value of the consistency factor for diagnosing local demagnetization faults. This represents the normalized value of the amplitude at the offset frequency in the spectrum of the zero-sequence voltage signal, which is also the normalized value of the characteristic value for local demagnetization fault diagnosis. A schematic diagram of the process for obtaining the comprehensive diagnostic characteristic value for local demagnetization fault diagnosis is shown below. Figure 4 As shown.

[0105] In this embodiment, during the calculation of the comprehensive diagnostic feature value, the normalized values ​​of the feature quantities, the normalized values ​​of the consistency factors, and the normalized values ​​of the feature values ​​are all obtained using the Min-Max normalization method. The Min-Max normalization method is a well-known technique and will not be described in detail here.

[0106] This application integrates phase current analysis method and zero-sequence voltage response analysis method based on high-frequency current signal injection to extract features of high-frequency electrical signals of early faults in three-phase permanent magnet synchronous motors in EMB systems. It can perform accurate fault diagnosis under various operating conditions, enhance the reliability of fault identification, help to detect potential faults in three-phase permanent magnet synchronous motors in EMB systems in the early stage of faults, and improve driving safety.

[0107] Based on the same inventive concept as the above method, this application embodiment also provides a high-frequency electrical signal feature extraction system for an EMB system, 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-described high-frequency electrical signal feature extraction methods for an EMB system.

[0108] In summary, this application considers that inter-turn short-circuit faults can cause an imbalance in the waveform of current data between the faulty and non-faulty phases. By dividing the current data into different periods and comparing the distribution differences of current data in different phases within different periods, this waveform imbalance can be accurately captured, thus obtaining a characteristic quantity that can effectively characterize inter-turn short-circuit faults. Local demagnetization faults can cause the current data of each phase of a three-phase permanent magnet synchronous motor to exhibit certain similarities and discrete changes in peak-valley differences. By obtaining the peak-valley difference coefficient, characteristic information related to local demagnetization faults can be extracted from the variation pattern of current data, making local demagnetization faults distinguishable from other faults in terms of characteristics, thereby enhancing the ability to differentiate faults in fault diagnosis.

[0109] Furthermore, by assessing the consistency of inter-turn short-circuit faults and local demagnetization faults within different cycles, the reliability of the extracted feature quantities of inter-turn short-circuit faults and local demagnetization faults is evaluated. Under complex operating conditions such as low speed, light load, and high noise, it can effectively filter false alarms caused by various interference factors, thereby enhancing the robustness of feature extraction and fault diagnosis of high-frequency electrical signals of three-phase permanent magnet synchronous motors in the EMB system.

[0110] Furthermore, by fusing phase current analysis with zero-sequence voltage response analysis based on high-frequency current signal injection, feature extraction of high-frequency electrical signals of early faults in three-phase permanent magnet synchronous motors of EMB systems can be performed. This allows for the fusion of various fault-related information, enabling the effective extraction of fault-related feature information under various operating conditions. This helps to detect potential faults in three-phase permanent magnet synchronous motors of EMB systems in their early stages, providing strong support for timely measures to ensure driving safety.

[0111] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in 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 description, and 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 reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0112] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from its essential characteristics. Therefore, the embodiments described above should be considered exemplary and non-limiting in all respects.

Claims

1. A method for extracting high-frequency electrical signal features for an EMB system, characterized in that, The method includes the following steps: Real-time acquisition of phase current data of the three-phase permanent magnet synchronous motor of the EMB system during the first preset time period, and zero-sequence voltage signal of the three-phase permanent magnet synchronous motor after injecting high-frequency current signal into the stator winding of the three-phase permanent magnet synchronous motor during the second preset time period. The first time period is divided into cycles. By comparing the distribution differences of current data of different phases in each cycle, the waveform imbalance of each cycle is obtained, and then the characteristic quantity for inter-turn short circuit fault diagnosis is obtained. The peak-valley difference of current data in each phase within each cycle is measured. By measuring the similarity of the peak-valley difference of different phases within each cycle and the dispersion of the peak-valley difference of each phase within each cycle, the peak-valley difference coefficient of each cycle is obtained, and then the characteristic quantity for local demagnetization fault diagnosis is obtained. By analyzing the discreteness of waveform imbalance and peak-valley difference coefficients for all cycles, the consistency factors for inter-turn short-circuit fault diagnosis and local demagnetization fault diagnosis are obtained. By statistically analyzing the injection frequency of the high-frequency current signal and the offset frequency when a local demagnetization fault occurs under the current operating conditions, combined with the amplitude value in the spectrum of the zero-sequence voltage signal, and the characteristic quantities and consistency factors of inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis, comprehensive diagnostic characteristic values ​​for inter-turn short circuit fault diagnosis and local demagnetization fault diagnosis are obtained respectively. The process of obtaining the waveform imbalance is as follows: Obtain the peak and trough points of the current data of each phase in each cycle in the time sequence, and calculate the difference between each peak point and its adjacent trough point; take the average of all the corresponding difference values ​​in each cycle as the average peak-trough distance of each phase in each cycle. Arrange the average peak-to-valley distances of all phases within each period in descending order. The expression for the waveform imbalance of each period is: In the formula, This represents the waveform imbalance in the i-th cycle; This represents the maximum value among the average peak-to-valley distances of all phases within the i-th period; This represents the median value of the average peak-to-valley distance among all phases within the i-th period; This represents the minimum average peak-to-valley distance among all phases within the i-th period; Indicates a preset positive number; The process for obtaining the peak-valley difference coefficient is as follows: The similarity of the out-of-phase waveforms in each period is obtained by measuring the similarity of the difference values ​​between different phases within each period. The volatility of the in-phase waveform in each period is obtained by measuring the degree of dispersion of the corresponding difference values ​​within each period. The similarity of the out-of-phase waveform and the volatility of the in-phase waveform are respectively mapped to a first positive number and a second positive number; The peak-valley difference coefficient is the product of the first positive number and the second positive number.

2. The method for extracting high-frequency electrical signal features for an EMB system as described in claim 1, characterized in that, The characteristic quantity for inter-turn short-circuit fault diagnosis is the average value of waveform imbalance over all cycles.

3. The method for extracting high-frequency electrical signal features for an EMB system as described in claim 1, characterized in that, The process of obtaining the similarity of the out-of-phase waveforms is as follows: Calculate the similarity of the difference values ​​between any two phases in each period, wherein the similarity of the out-of-phase waveforms is the mean of the similarities between all any two phases in each period.

4. The high-frequency electrical signal feature extraction method for an EMB system as described in claim 1, characterized in that, The process for obtaining the in-phase waveform fluctuation is as follows: The dispersion of all the difference values ​​of each phase within each period is denoted as the peak-valley distance dispersion of each phase within each period; The in-phase waveform volatility is the average of the peak-to-valley distance dispersion of all phases within each period.

5. The high-frequency electrical signal feature extraction method for an EMB system as described in claim 1, characterized in that, The characteristic quantity for diagnosing local demagnetization faults is the average of the peak-valley difference coefficients for all cycles.

6. The high-frequency electrical signal feature extraction method for an EMB system as described in claim 1, characterized in that, The method for obtaining the consistency factor for inter-turn short-circuit fault diagnosis and local demagnetization fault diagnosis is as follows: The consistency factor for inter-turn short-circuit fault diagnosis is the reciprocal of the sum of the dispersion of waveform imbalance in all cycles and a preset constant greater than 0. The consistency factor for the diagnosis of local demagnetization faults is the reciprocal of the sum of the dispersion of the peak-valley difference coefficients of all periods and a preset constant greater than 0.

7. The high-frequency electrical signal feature extraction method for an EMB system as described in claim 1, characterized in that, The expressions for the combined diagnostic feature values ​​of the inter-turn short-circuit fault diagnosis and the local demagnetization fault diagnosis are as follows: In the formula, This represents the comprehensive diagnostic characteristic value for inter-turn short-circuit fault diagnosis; Normalized values ​​of characteristic quantities representing inter-turn short-circuit fault diagnosis; This represents the normalized value of the consistency factor for inter-turn short-circuit fault diagnosis. The normalized value of the amplitude at the injection frequency in the spectrum of the zero-sequence voltage signal; In the formula, The comprehensive diagnostic feature value represents the diagnosis of local demagnetization faults; Normalized values ​​of characteristic quantities representing local demagnetization fault diagnosis; This represents the normalized value of the consistency factor for diagnosing local demagnetization faults. This represents the normalized value of the amplitude at the offset frequency in the spectrum of the zero-sequence voltage signal.

8. A high-frequency electrical signal feature extraction system for an EMB system, 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, it implements the steps of the high-frequency electrical signal feature extraction method for an EMB system as described in any one of claims 1-7.

Images