High-side driving circuit fault diagnosis method and system

Abstract

The invention relates to the technical field of circuit fault diagnosis, in particular to a high-side drive circuit fault diagnosis method and system, and the method comprises the steps: firstly processing collected power data through employing an unsupervised anomaly detection algorithm, and obtaining an initial anomaly score; generating a correction factor by calculating a power fluctuation factor reflecting that the current power deviates from the historical average level and combining the power fluctuation accumulation degree in a future time window; correcting the initial abnormal score by using a correction factor to obtain a target abnormal score; and judging a fault according to whether the target abnormal score meets a preset condition. According to the method, a statistical model and physical characteristic analysis are combined, noise interference is effectively suppressed, and the accuracy and reliability of fault diagnosis are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of circuit fault diagnosis, and in particular to a high-side drive circuit fault diagnosis method and system. BACKGROUND

[0002] The high-side drive circuit is a key component in power electronic systems and control systems, and its stable operation is crucial to the safety and performance of the power electronic system. Therefore, developing an efficient and accurate online fault diagnosis method to timely detect and locate potential faults in the circuit has important engineering application value.

[0003] At present, methods based on circuit operation data for intelligent fault diagnosis have gradually attracted attention. Unsupervised anomaly detection algorithms represented by Isolation Forest have been applied in practice due to their characteristics of not requiring pre-labeled fault samples and being able to automatically discover abnormal patterns from data. The Isolation Forest algorithm isolates data points by building random trees and determines whether a data point is abnormal according to the average path length required to isolate it; generally, abnormal points are more easily isolated due to their rarity and unique characteristics, so the path length is shorter.

[0004] However, when applied to high-side drive circuits and other scenarios with strong electromagnetic interference and complex working condition changes, such algorithms that rely purely on data statistical distribution expose their inherent defects: the Isolation Forest algorithm can effectively identify outliers in data streams, but cannot distinguish between transient high-amplitude noise glitches and persistent power anomalies representing real faults in a physical sense. This leads to a high false positive rate when facing complex noise interference, affecting the reliability and practicality of the diagnosis results. SUMMARY

[0005] To solve the technical problem that unsupervised anomaly detection algorithms such as Isolation Forest cannot distinguish between transient high-amplitude noise and real fault persistent power anomalies, and thus easily misjudge normal transient disturbances as faults, the present application provides solutions in the following aspects.

[0006] In a first aspect, the present application provides a high-side drive circuit fault diagnosis method, which comprises the steps of: Power data of the high-side drive circuit is collected at a preset sampling frequency for several time periods. An unsupervised anomaly detection algorithm is used to process the power data at each sampling time to obtain an initial anomaly score for each sampling time. For any given sampling time, a power fluctuation factor is calculated based on the power data at that sampling time and the total energy transmitted by the circuit within its historical neighborhood window. A correction factor is calculated for that sampling time. The correction factor is positively correlated with both the power fluctuation factor at that sampling time and the power fluctuation degree of the future neighborhood window at that sampling time. The initial anomaly score is corrected using the correction factor at that sampling time to obtain a target anomaly score. The target anomaly score is positively correlated with both the initial anomaly score at that sampling time and its correction factor. The target anomaly score is calculated for each sampling time. If the target anomaly score meets a preset judgment condition, a high-side drive circuit fault is determined.

[0007] This invention first obtains an initial anomaly score using an unsupervised algorithm, then introduces a power fluctuation factor and a future neighborhood window power fluctuation degree. The former quantifies the degree of instantaneous power mutation, while the latter characterizes the persistence of anomalies. Based on these two factors, a correction factor is constructed to dynamically correct the initial anomaly score. This dual verification mechanism, combining statistical models and physical characteristics, can distinguish between two types of signals: one is a real fault with a long duration and significant energy anomaly, and the other is a noise spike with a short duration and low total energy. Ultimately, while ensuring high sensitivity to real faults, it effectively reduces the false alarm rate and improves the accuracy of the diagnostic system.

[0008] Preferably, the step of calculating the power fluctuation factor based on the power data at the sampling time and the total energy transmitted by the circuit within its historical neighborhood window includes: obtaining the power data at the sampling time and the average power value within its historical neighborhood window; and performing normalization processing based on the deviation between the power data at the sampling time and the average power data, combined with the rated power value of the circuit, to obtain the power fluctuation factor.

[0009] This invention standardizes the measurement of power fluctuations by calculating the deviation between the current power value and the historical average power and performing normalization processing, which provides stable and reliable basic data for subsequent calculation of correction factors.

[0010] Preferably, the power fluctuation factor satisfies the following relationship: ; in, It is the first Power fluctuation factor at the sampling time; It is the first Power data at the sampling time, It is the length of the historical neighborhood window. It is the first Within the historical neighborhood window of the sampling time Power data at any given time; It is the absolute value symbol; This is the rated power value of the circuit; It is the index value at the sampling time.

[0011] This invention quantifies the degree of power fluctuation by calculating the absolute deviation between the current sampling power value and the average power within a historical neighborhood window, and then normalizing it using the circuit's rated power. On one hand, the absolute value calculation ensures that deviations in both positive and negative directions are considered, avoiding misjudgments caused by offsetting deviations. On the other hand, the normalization operation by dividing by the rated power eliminates differences between circuits of different power levels, providing a unified measurement scale for the fluctuation factor, making it applicable to high-side drive circuits of different specifications. Simultaneously, the introduction of the historical neighborhood window allows the assessment of current power fluctuations to be based on a reference to recent stable states, improving the accuracy of identifying abnormal fluctuations and providing standardized basic parameters for subsequent correction factor calculations.

[0012] Preferably, obtaining the power fluctuation level of the future neighborhood window at the sampling time includes: obtaining the average value of the power data within the future neighborhood window; calculating the absolute value of the difference between each power data within the future neighborhood window and the average value; and integrating the absolute value within the future neighborhood window to obtain the power fluctuation level.

[0013] This invention effectively quantifies the sustained fluctuation characteristics after an anomaly occurs by first calculating the average power within a future neighborhood window, then calculating and integrating the absolute deviation of the power from the average value at each time point. This method establishes a baseline state within the window through the average value, and accumulates the total fluctuation within the entire window through absolute deviation and integration, thus fully reflecting the persistence and severity of the anomaly. This calculation method can effectively distinguish between instantaneous noise and genuine faults.

[0014] Preferably, the correction factor satisfies the following relationship: ; in, It is the first Correction factor at sampling time; It is the first Power fluctuation factor at the sampling time; It is the index value at the sampling time; It is the length of the future neighborhood window; It is the first Within the future neighborhood window at the sampling time, the first Power data at any given time; It is the first an average power value in a future neighborhood window of the sampling time; is an absolute value symbol; is a standard normalization function.

[0015] The application can capture the severity of the instantaneous mutation and take into account the duration of the abnormal state by multiplying the power fluctuation factor reflecting the current instantaneous mutation degree and the future neighborhood window power fluctuation integral term embodying the persistent influence of the abnormality, thereby avoiding one-sidedness of single-dimensional evaluation.

[0016] Preferably, the step of correcting the initial abnormality score of the sampling time by using the correction factor thereof to obtain a target abnormality score comprises: multiplying the initial abnormality score of the sampling time by the correction factor thereof to obtain the target abnormality score.

[0017] Preferably, the preset determination condition is that the target abnormality score is greater than a preset fault threshold.

[0018] Preferably, the unsupervised anomaly detection algorithm is an isolation forest algorithm.

[0019] Preferably, the length of the historical neighborhood window is the same as that of the future neighborhood window.

[0020] In a second aspect, the application provides a high-side drive circuit fault diagnosis system, comprising a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the first aspect of the application is implemented.

[0021] By using the above technical solution, the first aspect of the application is generated into a computer program and stored in the memory to be loaded and executed by the processor, so that a terminal device is manufactured according to the memory and the processor, and the use is facilitated.

[0022] The application has the following beneficial effects: the application firstly uses an unsupervised anomaly detection algorithm to process collected power data to obtain an initial abnormality score. Then, a correction mechanism based on physical characteristics is introduced: a power fluctuation factor reflecting the current power deviation from the historical average level is calculated, and the power fluctuation accumulation degree in the future neighborhood window is combined to generate a correction factor. The correction factor is used to correct the initial abnormality score to obtain a target abnormality score that can better reflect the real fault possibility. Finally, whether the target abnormality score meets a preset condition is determined to determine the fault. The application combines statistical models and physical characteristic analysis, effectively suppresses noise interference, and improves the accuracy and reliability of fault diagnosis. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1A flow chart of a high-side drive circuit fault diagnosis method provided by an embodiment of the present application is shown in the figure; Figure 2 A curve graph of original power data of a high-side drive circuit provided by an embodiment of the present application is shown in the figure; Figure 3 A curve graph of initial abnormal score of each sampling time provided by an embodiment of the present application is shown in the figure; Figure 4 A curve graph of target abnormal score of each sampling time provided by an embodiment of the present application is shown in the figure; Figure 5 A structure block diagram of a high-side drive circuit fault diagnosis system provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0024] The first aspect of the embodiment of the present application provides a high-side drive circuit fault diagnosis method, as shown in the figure, the method comprises steps S100-S600: Figure 1 Step S100, collecting power data of the high-side drive circuit at several time points according to a preset sampling frequency. Step S100, collecting power data of the high-side drive circuit at several time points according to a preset sampling frequency.

[0025] It should be noted that the power signal of the high-side drive circuit can directly reflect its working state, when the circuit has a short circuit, open circuit or component aging fault, the power consumption will show a feature significantly different from the normal state. Therefore, collecting accurate power data is the premise of subsequent accurate fault diagnosis.

[0026] Specifically, the power sensor is deployed at the key node of the target high-side drive circuit, and the power input end or the output end can be selected; among them, the power input end can collect the overall power consumption of the circuit, reflecting the matching state of global power supply and load; the output end can collect the load side power, which is directly related to the fault characteristics of the drive object. The real-time power data in the circuit running process is continuously collected at a preset sampling frequency, and finally a power data time sequence containing time dimension is formed.

[0027] As a preferred embodiment, the preset sampling frequency can be set to 1kHz; the selection of this parameter is based on the balance between accuracy and cost: too low sampling frequency may miss the characteristics of transient faults such as instantaneous short circuit, which usually lasts more than 1ms; too high frequency will greatly increase the data processing burden and storage cost. The 1kHz sampling frequency corresponds to a 1ms sampling interval, which can not only cover the feature capture needs of most conventional instantaneous faults, but also avoid unnecessary resource consumption.

[0028] At this point, the high-side drive circuit power data containing several sampling time points is obtained.

[0029] Step S200, processing the power data of each sampling time using an unsupervised anomaly detection algorithm to obtain an initial anomaly score of each sampling time.

[0030] It should be noted that after obtaining the original power data, the data needs to be preliminarily screened to identify potential abnormal points. The unsupervised anomaly detection algorithm is very suitable for online real-time monitoring scenarios because it does not need to pre-label fault samples and can directly find anomalies from the distribution characteristics of the data itself, so this type of algorithm is selected to complete the data processing of this step.

[0031] In this embodiment, the unsupervised anomaly detection algorithm is preferably an isolation forest algorithm; the isolation forest algorithm isolates data points by randomly building multiple decision trees, i.e., forests. The basic idea is that abnormal points have rare and different characteristics, and are usually closer to the root node in the decision tree than normal points, i.e., they can be isolated by a shorter path. When the algorithm runs, it calculates the average path length required to isolate each data point, which quantifies the abnormality of the data point, and finally obtains the initial anomaly score. As shown in FIG. 2, Figures 2-3 Figure 2 is a curve graph of the original power data of the high-side drive circuit, the horizontal axis represents time, in seconds, and the vertical axis represents power, in watts, Figure 2 shows the change of the original power data collected at each sampling time. Figure 3 is a graph of the initial anomaly score of each sampling time after processing by the isolation forest algorithm. The horizontal axis also represents time, in seconds, and the vertical axis represents the initial anomaly score, which reflects the abnormality of the data at each time after processing by the isolation forest algorithm.

[0032] As a feasible implementation, the number of trees of the isolation forest algorithm can be set to 100, and the number of sub-samples of each tree can be set to 256. These hyperparameters can be empirically set according to the data characteristics of the actual application scenario, or can be optimized through cross-validation to adapt to the anomaly detection needs in different scenarios.

[0033] In addition, other unsupervised anomaly detection algorithms can also be applied to the present application to obtain the initial anomaly score, such as the local anomaly factor, the one-class support vector machine, etc. The above unsupervised anomaly detection algorithms are prior art and will not be described in detail here.

[0034] At this point, the initial anomaly score of each sampling time is obtained.

[0035] Step S300, for any sampling time, calculating a power fluctuation factor according to the power data of the sampling time and the total energy transmitted by the circuit in the historical neighborhood window thereof.

[0036] ​It should be noted that the initial anomaly score is only judged from a statistical point of view, and the dynamic change characteristics of the circuit power data in the time dimension are not considered. The real circuit fault is usually not isolated, but related to the state of the previous moment. Therefore, the power fluctuation factor is introduced to quantify the deviation of the current power value relative to its immediate historical state, and the abnormal judgment basis of the dynamic dimension is supplemented.

[0037] Specifically, for the first sampling moment, first define a length of historical neighborhood window, which contains all power data from the first sampling moment to the first sampling moment, that is, the window covers consecutive historical sampling points, and does not include the first sampling moment itself. Among them, the window length should be determined in combination with the running period characteristics of the circuit and the sampling frequency. The setting principle is to cover the typical stable fluctuation interval of the normal operation of the circuit, and to avoid the sensitivity of recent abnormalities due to too long window, or misjudgment of normal high-frequency fluctuations due to too short window.

[0038] For example, the sampling frequency is 1kHz, that is, 1 data point is collected every 1ms, and the power stable period of the high-side drive circuit during normal operation is 50ms-100ms, then can be set to 50-100 preferentially; if the circuit load is high-frequency fluctuation type, the normal fluctuation period is as short as 10ms-20ms, then can be adjusted to 10-20 to ensure that normal high-frequency fluctuations and abnormal mutations can be distinguished.

[0039] During the operation of the circuit, the greater the deviation of the power value at the current moment from the average value of the power data in the historical neighborhood window, the more intense the power fluctuation at the current moment, and the more likely it is to be accompanied by an abnormal state.

[0040] According to the above logic, the power fluctuation factor satisfies the relationship: ; Among them, is the power fluctuation factor at the first sampling moment; is the power data at the first sampling moment, is the length of the historical neighborhood window, is the power data at the first sampling moment in the historical neighborhood window at the first sampling moment; is the absolute value symbol; is the rated power value of the circuit; is the index value of the sampling time.

[0041] In the formula, represents the total energy transmitted by the circuit in the historical neighborhood window of the time, and the value multiplied by represents the average power in the historical neighborhood window of the time. When the power value at the time deviates significantly from the average power of its previous times, the value of the molecule will increase, thereby causing the value of the power fluctuation factor to also increase, indicating that a more severe power mutation occurs at the sampling time, and such mutations are important indicators of potential circuit abnormalities.

[0042] At this point, the power fluctuation factor corresponding to each sampling time is obtained.

[0043] Step S400, calculate the correction factor of the sampling time.

[0044] It should be noted that the core of this step is to distinguish the physical characteristics of real faults and transient noise. Real circuit faults are usually accompanied by continuous abnormal energy dissipation; while noise pulses, although the instantaneous power can be high, the duration is extremely short, and the overall abnormal energy scale is limited. Therefore, by analyzing the energy scale characteristics in a short period of time after the occurrence of an abnormal point, a correction factor can be constructed to adjust the credibility of the initial abnormal score.

[0045] Specifically, for the sampling time, a future neighborhood window with a length of is defined again, which contains power data from to time. The length of the future neighborhood window and the length of the historical neighborhood window need to be consistent, keeping consistent with , which can ensure that the stable benchmark of historical data and the continuous fluctuation of future data are unified in the time dimension, avoiding the imbalance of historical fluctuation judgment and future duration judgment due to the difference in window length, and improving the objectivity of the correction factor.

[0046] The severity of an abnormal event depends not only on the instantaneous impact strength at the time of its occurrence, but also on the subsequent induced continuous power disturbance. Therefore, the correction factor should be positively correlated with the power fluctuation factor at the current time and the power fluctuation degree of the future neighborhood window. The power fluctuation degree of the future neighborhood window can be quantified by calculating the cumulative deviation of the power at each time from the average power of the window.

[0047] ​Based on the above logic, the correction factor satisfies the following relation: ; in, It is the first Correction factor at sampling time; It is the first Power fluctuation factor at the sampling time; It is the index value at the sampling time; It is the length of the future neighborhood window; It is the first Within the future neighborhood window at the sampling time, the first Power data at any given time; It is the first The average power value within a future neighborhood window at the sampling time; It is the absolute value symbol; These are standard normalization functions, such as min-max normalization, used to map computation results to... The interval is used to facilitate subsequent processing.

[0048] This relation consists of two parts, the first part... Reflects the degree of change at the current moment; Part Two This reflects the degree of sustained power fluctuation over a future period. When transient noise occurs, although... The value may be large, but because the power will quickly return to normal, the fluctuation within the future neighborhood window will be small, resulting in a small final correction factor. Conversely, when a real fault occurs, The impact will be significant, and due to the persistence of the fault, the power within the neighborhood window will continue to fluctuate abnormally in the future, causing... It is also very large, which leads to a significant increase in the final correction factor.

[0049] At this point, the correction factors for each sampling time have been obtained.

[0050] Step S500: Correct the initial anomaly score using the correction factor at the sampling time to obtain the target anomaly score.

[0051] It should be noted that this step aims to integrate the dynamic characteristics of the time series with the static statistical characteristics of the data, thereby obtaining a more reliable comprehensive evaluation index that better reflects the actual probability of failure.

[0052] Specifically, the target anomaly score is obtained by directly multiplying the initial anomaly score corresponding to the sampling time by the correction factor. The core function of this calculation logic is: when a persistent anomaly is detected, the correction factor... The larger, the initial abnormal score can be amplified, so that the final target abnormal score is more likely to exceed the fault determination threshold, ensuring that real faults are not missed; on the contrary, when only transient disturbances such as noise are detected, the correction factor The smaller, the initial abnormal score can be suppressed, so that it is lower than the fault determination threshold, effectively avoiding false positives caused by noise.

[0053] As Figure 4 shown, it is a target abnormal score graph after correction of each sampling time, wherein the horizontal axis represents time, unit: second; the vertical axis represents the target abnormal score, which is used to measure the abnormality of the data at each time. Compared with Figure 3 It can be seen that the target abnormal score obtained after the correction factor adjustment can more accurately identify the abnormality corresponding to the real fault, reduce the misjudgment of normal disturbances such as transient high-amplitude noise, and make the fault detection result more reliable.

[0054] At this point, the target abnormal score of each sampling time is obtained.

[0055] Step S600, the target abnormal score of each sampling time is calculated one by one, and the high-side drive circuit fault is determined in response to the target abnormal score meeting a preset determination condition.

[0056] It should be noted that this step is the final decision-making link of fault diagnosis. In this step, the target abnormal score which can comprehensively reflect the abnormality and continuity is compared with a preset threshold to realize the final determination of the circuit state.

[0057] Specifically, first, the preset determination condition is determined: when the target abnormal score of a certain sampling time is greater than the preset fault threshold, the fault determination condition is met. Based on this, the target abnormal scores of all sampling times obtained are determined one by one: the target abnormal score of each sampling time is compared with the fault threshold, and if the target abnormal score of a certain sampling time is greater than the threshold, it is determined that the high-side drive circuit has a fault at that time.

[0058] The setting of the fault threshold can be determined by offline test analysis on a large number of historical normal and fault data, aiming to achieve the best balance between the false negative rate and the false positive rate.

[0059] For example, in order to determine the fault threshold, 1000 normal samples and 100 known fault samples are collected, and their target abnormal scores are calculated. It is found that the scores of normal samples are concentrated in the interval of 0.1-0.5, while the scores of fault samples are all higher than 0.8. In order to balance the false negative rate and the false positive rate, we set the fault threshold to 0.85.

[0060] Once the fault is determined, the system can immediately generate a warning signal to inform the operation and maintenance personnel to carry out maintenance, thereby ensuring the safe and stable operation of the power electronic system.

[0061] The second aspect of the embodiment provides a high-side drive circuit fault diagnosis system, which comprises a memory and a processor, and the memory stores computer program instructions. Figure 5 As shown in the figure, the high-side drive circuit fault diagnosis system comprises a memory and a processor, and the memory stores computer program instructions.

[0062] The high-side drive circuit fault diagnosis system further comprises a communication bus and a communication interface and other components familiar to those skilled in the art, the settings and functions of which are known in the art, and thus will not be described here.

[0063] In the present application, the aforementioned memory can be any tangible medium containing or storing a program, which can be used by or in conjunction with an instruction execution system, device or apparatus. For example, the computer readable storage medium can be any suitable magnetic storage medium or magneto-optical storage medium, such as resistive random access memory, dynamic random access memory, static random access memory, enhanced dynamic random access memory, high bandwidth memory, hybrid memory cube, etc., or any other medium that can be used to store the required information and can be accessed by an application program, module or both. Any such computer storage medium can be part of the device or accessible or connectable to the device.

[0064] The above are the preferred embodiments of the present application, which do not limit the protection scope of the present application, therefore: any equivalent changes made in accordance with the structure, shape, principle of the present application should be covered within the protection scope of the present application.

Claims

1. A high-side driver circuit failure diagnosis method characterized by, The method comprises the steps of: collecting power data of the high-side drive circuit at several time points according to a preset sampling frequency; processing the power data at each sampling time point using an unsupervised anomaly detection algorithm to obtain an initial anomaly score at each sampling time point; for any sampling time point, calculating a power fluctuation factor according to the power data at the sampling time point and the total energy transmitted by the circuit within a historical neighborhood window of the sampling time point; calculating a correction factor of the sampling time point; the correction factor is positively correlated with the power fluctuation factor of the sampling time point and the power fluctuation degree of a future neighborhood window of the sampling time point; correcting the initial anomaly score of the sampling time point using the correction factor to obtain a target anomaly score; the target anomaly score is positively correlated with the initial anomaly score of the sampling time point and the correction factor thereof; calculating the target anomaly score of all sampling time points one by one, and determining a high-side drive circuit fault in response to the target anomaly score satisfying a preset determination condition.

2. The high-side driver circuit fault diagnosis method of claim 1, wherein The calculation of the power fluctuation factor according to the power data at the sampling time point and the total energy transmitted by the circuit within the historical neighborhood window of the sampling time point comprises: obtaining the power data at the sampling time point and the average power value within the historical neighborhood window thereof; normalizing the deviation degree of the power data at the sampling time point from the average power data in combination with the rated power value of the circuit to obtain the power fluctuation factor.

3. The high-side driver circuit fault diagnosis method according to claim 1 or 2, characterized in that, The power fluctuation factor satisfies the relationship: ； in, It is the first Power fluctuation factor at the sampling time; It is the first Power data at the sampling time, It is the length of the historical neighborhood window. It is the first Within the historical neighborhood window of the sampling time Power data at any given time; It is the absolute value symbol; This is the rated power value of the circuit; It is the index value at the sampling time.

4. The high-side driver circuit fault diagnosis method of claim 1, wherein The acquisition of the power fluctuation degree of the future neighborhood window of the sampling time point comprises: obtaining the average value of the power data within the future neighborhood window; calculating the absolute value of the difference between each power data within the future neighborhood window and the average value; integrating the absolute value within the future neighborhood window to obtain the power fluctuation degree.

5. The high-side driver circuit fault diagnosis method of claim 1, wherein The correction factor satisfies the relationship: ； wherein, is the first is a correction factor for the sampling instant; is the first is a power fluctuation factor for the sampling instant; is an index value for the sampling instant; is the length of the future neighborhood window; is the first is the power data at the instant within the future neighborhood window for the sampling instant; is the average power value within the future neighborhood window for the sampling instant; is the average power value within the future neighborhood window for the sampling instant; is the absolute value sign; is the standard normalization function.

6. The high-side driver circuit fault diagnosis method of claim 1, wherein The correction of the initial anomaly score of the sampling time point using the correction factor to obtain the target anomaly score comprises: multiplying the initial anomaly score of the sampling time point by the correction factor to obtain the target anomaly score.

7. The high-side driver circuit fault diagnosis method of claim 1, wherein The preset determination condition is that the target anomaly score is greater than a preset fault threshold.

8. The high-side driver circuit fault diagnosis method of claim 1, wherein The unsupervised anomaly detection algorithm is an isolation forest algorithm.

9. The high-side driver circuit fault diagnosis method of claim 1, wherein The lengths of the historical neighborhood window and the future neighborhood window are the same.

10. A high-side driver circuit fault diagnosis system characterized by comprising: The high-side drive circuit fault diagnosis system comprises a processor and a memory, and the memory stores computer program instructions, which, when executed by the processor, implement a high-side drive circuit fault diagnosis method according to any one of claims 1-9.

Images