A fault diagnosis method for an electric vehicle motor drive control system

Abstract

This application discloses a fault diagnosis method for an electric vehicle motor drive control system, belonging to the field of fault diagnosis technology. The method includes: globally acquiring real-time operating data to obtain a global operating image; constructing a three-dimensional operating model based on the global operating image; restoring the electrical and mechanical coupling relationship of the three-dimensional operating model to obtain operating mechanism data; estimating the anomaly probability of the motor drive control system based on the operating mechanism data and the global operating image to obtain fault probability data; calculating the load impact energy increment from the fault probability data to obtain abnormal impact energy data; assigning local risk levels to the electric vehicle motor drive control system based on the fault probability data and abnormal impact energy data to form a local risk level map; constructing a safety hazard identification optimization model; and sending the safety hazard identification optimization model to an on-board diagnostic terminal. This application effectively improves fault diagnosis efficiency.

Description

Technical Field

[0001] This application relates to the field of fault diagnosis technology, and in particular to a fault diagnosis method for an electric vehicle motor drive control system. Background Technology

[0002] With the rapid development of the new energy vehicle industry, the reliability of the electric vehicle motor drive control system, as a core power unit, directly determines the vehicle's driving safety and range. Currently, electric vehicle motor drive systems are highly integrated and operate under complex conditions, making them prone to malfunctions due to aging electrical components, mechanical wear, and thermal management failures. Failure to diagnose these issues in a timely manner can lead to serious accidents such as motor burnout and power interruption.

[0003] In related technologies, traditional fault diagnosis methods mostly rely on manual inspection or single-point sensor monitoring, which suffers from data fragmentation and limited coverage dimensions, making it difficult to capture early hidden faults. Furthermore, they lack dynamic trend prediction and visualization modeling capabilities, failing to meet the needs of accurate and early diagnosis under the complex operating conditions of electric vehicles, thus reducing fault diagnosis efficiency and leaving room for improvement. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this application provides a fault diagnosis method for an electric vehicle motor drive control system.

[0005] In a first aspect, this application provides a fault diagnosis method for an electric vehicle motor drive control system, comprising the following steps:

[0006] Step S1: Acquire real-time operating data through multi-source sensor units deployed in the electric vehicle motor drive control system, and perform global acquisition of the real-time operating data. Obtain a global operating image based on the global acquisition results, and construct a three-dimensional operating model based on the global operating image.

[0007] Step S2: Reconstruct the electrical and mechanical coupling relationship of the three-dimensional operation model to obtain operation mechanism data; Based on the operation mechanism data and the global operation image, estimate the anomaly probability of the motor drive control system to obtain fault probability data; Calculate the load impact energy increment of the fault probability data to obtain abnormal impact energy data.

[0008] Step S3: Assign local risk level values ​​to the electric vehicle motor drive control system based on the fault probability data and the abnormal impact energy data to form a local risk level map;

[0009] Step S4: Construct a safety hazard identification and optimization model for the motor drive control system based on the local risk level map, and send the safety hazard identification and optimization model to the on-board diagnostic terminal to perform fault diagnosis of the electric vehicle motor drive control system.

[0010] Preferably, step S1 includes the following steps:

[0011] Step S11: By using sensor units deployed on the motor controller, motor stator, and motor shaft, the global operating data of the electric vehicle motor drive control system under different operating conditions is collected in real time to form a global operating image;

[0012] Step S12: Perform data normalization and distribution harmonic processing on the global running image, and obtain the running state harmonic image based on the processing result;

[0013] Step S13: Mark the running feature points on the running status harmonization image. The running feature points include motor current fluctuation points, voltage drop points, speed abnormal points, and temperature rise points, to obtain running feature point marking data.

[0014] Step S14: Based on the running state harmonized image and the running feature point marker data, construct a three-dimensional running model of the electric vehicle motor drive control system, and obtain a three-dimensional running model based on the construction results.

[0015] Preferably, step S2 includes the following steps:

[0016] Step S21: Perform anomaly analysis on the motor current and voltage trends of the three-dimensional operating model to obtain electrical characteristic trend data;

[0017] Step S22: Based on the electrical characteristic directional data, the electrical and mechanical coupling mechanism of the three-dimensional operation model is restored to obtain the electrical and mechanical operation mechanism data;

[0018] Step S23: Based on the electromechanical operation mechanism data and the global operation image, perform anomaly probability estimation on the electric vehicle motor drive control system to obtain fault probability data;

[0019] Step S24: Calculate the abnormal impact energy increment on the fault probability data to obtain abnormal impact energy data.

[0020] Preferably, step S23 includes the following steps:

[0021] Step S231: Perform vehicle power supply stability distribution analysis on the global operation image to obtain the power supply fluctuation distribution density;

[0022] Step S232: Based on the electromechanical operation mechanism data, decompose the power supply fluctuation distribution density and the three-dimensional operation model to obtain the electrical shock erosion degree data;

[0023] Step S233: Based on the electrical shock erosion degree data, perform a stability assessment on the motor controller to obtain critical support loss data;

[0024] Step S234: Perform risk difference coefficient deduction on the critical support loss data to obtain risk difference coefficient data;

[0025] Step S235: Based on the electrical shock erosion degree data, the critical support loss data, and the risk difference coefficient, the failure probability is estimated to obtain failure probability data.

[0026] Preferably, step S24 includes the following steps:

[0027] Step S241: Partition the fault probability data to obtain fault risk partition data;

[0028] Step S242: Calculate the current fluctuation difference between different risk areas based on the fault risk zoning data to obtain current fluctuation difference data;

[0029] Step S243: Calculate the temperature rise variance between different risk areas based on the risk zoning data to obtain temperature rise variance data;

[0030] Step S244: Based on the current fluctuation difference data and temperature rise variance data, perform load change energy simulation on the electromechanical operation mechanism data to obtain load change energy data;

[0031] Step S245: Calculate the abnormal impact energy increment based on the load change energy data to obtain abnormal impact energy data.

[0032] Preferably, step S244 includes the following steps:

[0033] Historical operating condition data is acquired, and the historical operating condition data is analyzed for spatiotemporal distribution to obtain load intensity distribution data;

[0034] Stability difference analysis was performed on the operating mechanism data of electrical machinery based on current fluctuation difference data and temperature rise variance data to obtain stability difference data.

[0035] The anti-interference capability is analyzed by the stability difference data to obtain anti-interference capability data;

[0036] Based on the load intensity distribution data, a multi-dimensional impact simulation evaluation is performed on the stability difference data and anti-interference capability data to obtain impact carrying density data.

[0037] Incremental analysis was performed on the impact-carrying density data to obtain incremental impact-carrying density data;

[0038] Based on the current fluctuation difference data and temperature rise variance data, load change energy simulation is performed on the impact-carrying density increment data to obtain load change energy data.

[0039] Preferably, step S3 includes the following steps:

[0040] Step S31: Normalize the abnormal impact energy data to obtain normalized abnormal impact energy data;

[0041] Step S32: Perform convolution calculation on the normalized abnormal impact energy data to obtain abnormal impact energy convolution data;

[0042] Step S33: Assign a local risk level to the electric vehicle motor drive control system based on the abnormal impact energy convolution data and the fault probability data to obtain a local risk level map.

[0043] Preferably, step S33 includes the following steps:

[0044] Step S331: Recalibrate the spatial coordinates of the three-dimensional running model to obtain the recalibrated spatial coordinates;

[0045] Step S332: Perform risk linkage analysis on the spatial recalibration coordinates based on the abnormal impact energy convolution data and fault probability data to obtain risk linkage data;

[0046] Step S333: Calculate the multivariate local risk index based on the abnormal impact energy convolution data, fault probability data, and risk linkage data to obtain the multivariate local risk index;

[0047] Step S334: Assign local risk level values ​​to the electric vehicle motor drive control system based on the multivariate local risk index to obtain a local risk level map.

[0048] Secondly, this application provides a fault diagnosis system for an electric vehicle motor drive control system, comprising:

[0049] The data acquisition module is used to acquire real-time operating data through multi-source sensor units deployed in the electric vehicle motor drive control system, and to perform global acquisition of the real-time operating data, obtain a global operating image based on the global acquisition results, and construct a three-dimensional operating model based on the global operating image.

[0050] The analysis and processing module is used to restore the electrical and mechanical coupling relationship of the three-dimensional operation model to obtain operation mechanism data; based on the operation mechanism data and the global operation image, the abnormal probability of the motor drive control system is estimated to obtain fault probability data; and the system load impact energy increment is calculated from the fault probability data to obtain abnormal impact energy data.

[0051] The risk classification module is used to assign local risk levels to the electric vehicle motor drive control system based on the fault probability data and the abnormal impact energy data, and form a local risk level map.

[0052] The optimization deployment module is used to construct a safety hazard identification optimization model for the motor drive control system based on the local risk level map, and send the safety hazard identification optimization model to the on-board diagnostic terminal to perform fault diagnosis of the electric vehicle motor drive control system.

[0053] Thirdly, this application provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform a fault diagnosis method for an electric vehicle motor drive control system as described in any of the above-mentioned claims.

[0054] In summary, this application includes at least one of the following beneficial technical effects:

[0055] This application provides a fault diagnosis method for an electric vehicle motor drive control system. It involves globally acquiring real-time operating data to obtain a global operating image, and constructing a three-dimensional operating model based on this image. The electrical and mechanical coupling relationships of the three-dimensional operating model are then reconstructed to obtain operating mechanism data. Based on the operating mechanism data and the global operating image, anomaly probability estimation is performed on the motor drive control system to obtain fault probability data. The load impact energy increment is calculated from the fault probability data to obtain abnormal impact energy data. Local risk levels are assigned to the electric vehicle motor drive control system based on the fault probability data and the abnormal impact energy data, forming a local risk level map. A safety hazard identification and optimization model is constructed and sent to an on-board diagnostic terminal, thereby effectively improving fault diagnosis efficiency. Attached Figure Description

[0056] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 This is a flowchart of a fault diagnosis method for an electric vehicle motor drive control system according to an embodiment of this application.

[0058] Figure 2 This is a schematic diagram of a fault diagnosis system for an electric vehicle motor drive control system according to an embodiment of this application. Detailed Implementation

[0059] The following is in conjunction with the appendix Figure 1-2 This application will be described in further detail.

[0060] Example 1

[0061] This application discloses a fault diagnosis method for an electric vehicle motor drive control system.

[0062] Reference Figure 1 A fault diagnosis method for an electric vehicle motor drive control system includes the following steps:

[0063] Step S1: Acquire real-time operating data through multi-source sensor units deployed in the electric vehicle motor drive control system, and perform global acquisition of the real-time operating data. Obtain a global operating image based on the global acquisition results, and construct a three-dimensional operating model based on the global operating image.

[0064] Step S2: Reconstruct the electrical and mechanical coupling relationship of the three-dimensional operation model to obtain operation mechanism data; Based on the operation mechanism data and the global operation image, estimate the anomaly probability of the motor drive control system to obtain fault probability data; Calculate the load impact energy increment of the fault probability data to obtain abnormal impact energy data.

[0065] Step S3: Assign local risk level values ​​to the electric vehicle motor drive control system based on the fault probability data and the abnormal impact energy data to form a local risk level map;

[0066] Step S4: Construct a safety hazard identification and optimization model for the motor drive control system based on the local risk level map, and send the safety hazard identification and optimization model to the on-board diagnostic terminal to perform fault diagnosis of the electric vehicle motor drive control system.

[0067] Specifically, multi-source sensor units are deployed in the core components of the electric vehicle motor drive control system (motor controller, stator winding, rotor shaft, cooling system interface, etc.). This includes: installing current and voltage sensors at the IGBT module of the motor controller to collect three-phase current and bus voltage data in real time; attaching temperature sensors to the surface of the stator core to record winding temperature rise; installing speed encoders and vibration sensors at the shaft end to capture speed fluctuations and radial vibration amplitude; and arranging flow and temperature sensors at the inlet and outlet of the cooling circuit to monitor the cooling medium status and globally collect real-time operating data. Parameters from different components and dimensions are integrated into a dataset with a unified time axis. A global operating image is generated using data visualization tools. Based on this global operating image, a three-dimensional operating model is constructed using 3D modeling technology. Using the physical dimensions of the motor drive system as a basis, real-time collected data such as temperature distribution and current density are attached to the corresponding components of the model using texture mapping technology.

[0068] Secondly, the electrical and mechanical coupling relationship of the three-dimensional operation model is restored: through mechanism analysis, a correlation model between electrical parameters and mechanical state is established. For example, the relationship between stator current harmonic content and shaft eccentricity is analyzed, the correlation between bus voltage fluctuation and motor controller capacitor aging is analyzed, and the mapping between vibration frequency and bearing wear degree is analyzed. Finally, the operation mechanism data is obtained. Based on the operation mechanism data and global operation image, a Bayesian probability model is used to estimate the probability of anomalies: the abnormal parameters in the global operation image are used as input, and combined with the prior probability of "abnormal parameter-fault type" in the operation mechanism data, the posterior probability of each component of the system failing is calculated to form fault probability data. The load impact energy increment is calculated on the fault probability data: combined with the current load of the motor, the energy change of high failure probability components when the load changes is analyzed. When the controller failure probability is high, the increase in load will lead to an increase in IGBT module power consumption, which will generate additional impact energy. By analyzing the energy difference before and after the load change, the abnormal impact energy data is obtained.

[0069] Next, local risk levels are assigned based on failure probability data and abnormal impact energy data: a risk assessment matrix is ​​set up, with failure probability (low: <30%, medium: 30%-60%, high: >60%) on the horizontal axis and abnormal impact energy (low: <50J, medium: 50-100J, high: >100J) on the vertical axis. The intersection of the two forms 9 risk levels (e.g., high probability + high energy is level 1 risk, low probability + low energy is level 9 risk). The 3D operating model is divided into local areas such as controller area, stator area, shaft area, and cooling area according to components. The failure probability and impact energy of each area are calculated one by one, and the risk level matrix is ​​matched to form a local risk level map with the risk level of each area marked.

[0070] Finally, a safety hazard identification optimization model is constructed based on the local risk level map: using the "region-risk level-abnormal parameter" data from the risk level map as training samples, the model is trained using a random forest algorithm, enabling it to automatically identify hazard types based on real-time collected parameters. Through transfer learning, historical fault data from different vehicle models and operating conditions are integrated to optimize the model's generalization ability, resulting in the optimized safety hazard identification model. This optimized model is then packaged into an executable program and sent to the vehicle diagnostic terminal via the vehicle Ethernet. The terminal calls the model in real-time to analyze sensor data, and when a hazard is identified, it immediately displays the fault location, risk level, and emergency suggestions, achieving automated and intelligent fault diagnosis.

[0071] It should be noted that step S1 includes the following steps:

[0072] Step S11: By using sensor units deployed on the motor controller, motor stator, and motor shaft, the global operating data of the electric vehicle motor drive control system under different operating conditions is collected in real time to form a global operating image;

[0073] Step S12: Perform data normalization and distribution harmonic processing on the global running image, and obtain the running state harmonic image based on the processing result;

[0074] Step S13: Mark the running feature points on the running status harmonization image. The running feature points include motor current fluctuation points, voltage drop points, speed abnormal points, and temperature rise points, to obtain running feature point marking data.

[0075] Step S14: Based on the running state harmonized image and the running feature point marker data, construct a three-dimensional running model of the electric vehicle motor drive control system, and obtain a three-dimensional running model based on the construction results.

[0076] Specifically, sensor units are precisely deployed for key components of the electric vehicle motor drive control system: Hall current sensors are installed at both ends of the IGBT bridge arm inside the motor controller to collect real-time U, V, and W phase currents; a voltage sensor is installed at the controller input bus to monitor the DC bus voltage; thermocouple temperature sensors are wound around the three-phase output terminals of the stator winding to record the real-time winding temperature; a vibration sensor is attached to the outside of the stator core to collect radial vibration acceleration; a photoelectric speed encoder is installed at the non-load end of the motor shaft to capture the shaft speed; a temperature sensor is installed at the shaft bearing housing to monitor the bearing temperature; and flow sensors and temperature sensors are installed at the inlet and outlet of the cooling system, respectively. The system records the flow rate and temperature difference of the cooling medium. The sensor unit is connected to the data acquisition module via the vehicle-mounted CAN-FD bus. The raw dataset corresponding to the global operating image is preprocessed: data normalization uses the min-max normalization algorithm to map parameters of different dimensions to the [0,1] interval, eliminating the interference of dimensional differences on subsequent analysis; distribution harmonic processing uses the 3σ criterion to identify outliers, calculates the mean μ and standard deviation σ of each parameter, removes data exceeding the range [μ-3σ,μ+3σ], and fills in missing data points using linear interpolation; high-frequency noise is eliminated through a data smoothing algorithm, ultimately obtaining a harmonic image of the operating state with uniform data distribution and no abnormal interference.

[0077] Operational feature points are marked on the harmonic image of the operating status: Based on the fault mechanism and safety threshold of the electric vehicle motor drive system, feature point identification rules are set: motor current fluctuation point is the moment when the normalized current value change rate exceeds 0.1 / s; voltage drop point is the moment when the bus voltage is lower than 95% of the rated value; abnormal speed point is the moment when the deviation between the actual speed and the target speed exceeds 5%; temperature rise point is the moment when the temperature change rate exceeds 2℃ / min. The harmonic image is automatically scanned by the feature recognition algorithm, and the time, parameter value and corresponding component of each feature point are recorded to form operational feature point marking data.

[0078] A 3D operating model is constructed based on the harmonic image of the operating status and the feature point marker data: Using the CAD 3D drawings of the motor drive system as the basic framework, the model is imported into SolidWorks software to recreate the physical structure of the controller, stator, shaft, and cooling pipes at a 1:1 scale. Real-time parameters from the harmonic image of the operating status are attached to the corresponding locations in the model using texture mapping technology. For example, the stator windings are assigned a color gradient based on temperature values, and the IGBT module of the controller is displayed with brightness based on current density. Abnormal points in the feature point marker data are marked with flashing red icons in the model and floating prompt boxes are added. The harmonic image and the 3D model are linked through a real-time data interface, allowing the model parameters to be dynamically updated with the real-time data of the harmonic image, ultimately resulting in a 3D operating model that dynamically reflects the system's operating status and abnormal characteristics.

[0079] Furthermore, step S2 includes the following steps:

[0080] Step S21: Perform anomaly analysis on the motor current and voltage trends of the three-dimensional operating model to obtain electrical characteristic trend data;

[0081] Step S22: Based on the electrical characteristic directional data, the electrical and mechanical coupling mechanism of the three-dimensional operation model is restored to obtain the electrical and mechanical operation mechanism data;

[0082] Step S23: Based on the electromechanical operation mechanism data and the global operation image, perform anomaly probability estimation on the electric vehicle motor drive control system to obtain fault probability data;

[0083] Step S24: Calculate the abnormal impact energy increment on the fault probability data to obtain abnormal impact energy data.

[0084] Specifically, anomaly analysis of motor current and voltage trends is performed on the 3D operating model: Five consecutive minutes of current and voltage data are extracted from the real-time data interface of the 3D model and imported into Matlab's signal analysis module. Fourier transform is used to decompose the harmonic components of the current signal, and wavelet transform is used to capture instantaneous fluctuations in the voltage signal. Time-amplitude curves and harmonic spectra of the current and voltage are plotted to analyze anomaly characteristics. For example, periodic peaks in the current curve may correspond to air gap inhomogeneity; continuous drops in the voltage curve may correspond to IGBT conduction anomalies in the controller. A trend fitting algorithm is used to analyze the long-term trends of current and voltage, recording the start time, amplitude, and associated components of the anomalies, forming electrical characteristic trend data including harmonic content, fluctuation frequency, and trend slope.

[0085] The electrical-mechanical coupling mechanism of the three-dimensional operation model is reconstructed based on electrical characteristic data: The correlation between electrical parameters and mechanical state is established through mechanism analysis. For example, according to motor theory, the stator current harmonic content is positively correlated with the air gap eccentricity. Combining the vibration data of the rotating shaft in the three-dimensional model, a coupling model of "current harmonic content - vibration amplitude - air gap eccentricity" is established. According to heat conduction theory, an increase in controller current leads to an increase in IGBT power consumption, which in turn increases the temperature. The increased temperature further increases the IGBT on-resistance, exacerbating current fluctuations. A positive feedback coupling model of "current fluctuation - temperature rise - resistance change" is established. According to tribological principles, shaft bearing wear leads to speed fluctuations, which in turn cause changes in the motor's back electromotive force, leading to bus voltage fluctuations. A coupling model of "bearing wear - speed fluctuation - voltage fluctuation" is established. These coupling relationships are quantified through mathematical models and input into the three-dimensional operation model, enabling the model to infer the mechanical state based on the electrical characteristic data, ultimately obtaining the electrical-mechanical operation mechanism data.

[0086] Anomaly probability estimation is performed by combining the operating mechanism data of electrical machinery with the global operating image: the coupling relationship between the real-time parameters in the global operating image and the operating mechanism data of electrical machinery is matched. For example, if the current harmonic content in the global image is 18%, and the mechanism data states that "the probability of air gap eccentricity fault is 85% corresponding to a harmonic content of 15%-20%", then the probability of air gap eccentricity fault is initially determined to be 85%. A multi-factor weighted algorithm is introduced to finally obtain the fault probability data.

[0087] Calculate the load impact energy increment based on the fault probability data: First, extract the current motor load data from the three-dimensional operating model and set the load change scenario; based on the electromechanical operating mechanism data, analyze the energy changes of high-fault-probability components under different loads. For example, when the stator air gap eccentricity fault probability is 85%, for every 10% increase in load rate, the stator iron loss increases by 5%. The increase in iron loss leads to increased energy loss, which in turn generates additional impact energy; analyze the energy difference before and after the load change using the principle of energy conservation. Impact energy increment = total energy after load increase - total energy under current load, where total energy includes electromagnetic energy, mechanical energy, and thermal energy; calculate the impact energy increment for different high-fault-probability components, summarize the abnormal impact energy data, and mark the correlation between energy increment and load change.

[0088] It should be noted that step S23 includes the following steps:

[0089] Step S231: Perform vehicle power supply stability distribution analysis on the global operation image to obtain the power supply fluctuation distribution density;

[0090] Step S232: Based on the electromechanical operation mechanism data, decompose the power supply fluctuation distribution density and the three-dimensional operation model to obtain the electrical shock erosion degree data;

[0091] Step S233: Based on the electrical shock erosion degree data, perform a stability assessment on the motor controller to obtain critical support loss data;

[0092] Step S234: Perform risk difference coefficient deduction on the critical support loss data to obtain risk difference coefficient data;

[0093] Step S235: Based on the electrical shock erosion degree data, the critical support loss data, and the risk difference coefficient, the failure probability is estimated to obtain failure probability data.

[0094] Specifically, a stability distribution analysis of the vehicle power supply is performed on the global operating image: One hour of continuous output voltage data from the vehicle power battery is extracted from the global operating image, and the voltage fluctuation value and frequency are calculated for each minute; statistical analysis methods are used to divide the voltage data into time intervals, calculate the mean and standard deviation of the fluctuation for each interval, and plot the voltage fluctuation time distribution curve; combined with the electric vehicle's driving conditions, the differences in power supply stability under different conditions are analyzed. For example, under congested conditions with frequent start-stop, the mean power supply fluctuation is 2V, and the fluctuation frequency is 15 times / minute; under high-speed conditions with a stable load, the mean fluctuation is 0.5V, and the fluctuation frequency is 3 times / minute; combining the mean fluctuation and frequency, the kernel density estimation method is used to calculate the power supply fluctuation distribution density, forming power supply fluctuation distribution density data, and labeling the operating conditions and times corresponding to the density peaks;

[0095] Based on the electrical machinery operation mechanism data, the distribution density of power supply fluctuations and the degree of electrical impact erosion in the three-dimensional operation model are decomposed: A correlation model of "power supply fluctuation - component damage" is extracted from the electrical machinery operation mechanism data. For example, the lifespan of the controller capacitor is inversely proportional to the square of the voltage fluctuation amplitude, and the degree of erosion of the stator winding insulation layer is positively correlated with the voltage fluctuation frequency. The power supply fluctuation distribution density data is input into this model, and combined with the material parameters of each component in the three-dimensional operation model, the degree of erosion of different components under the current fluctuation density is calculated. For example, in the distribution density range of 0.8 (congested working condition), the degree of erosion of the controller capacitor is 5% (capacitance value decays by 5%) within 1 hour; the degree of erosion of the stator insulation layer is 3% (insulation resistance decreases by 3%) in the distribution density range of 0.8. The impact process of power supply fluctuations on components is simulated using finite element analysis software (such as ANSYS), visually displaying the eroded areas (such as severe erosion at the capacitor pins and erosion of the stator winding end insulation layer), forming data on the degree of electrical impact erosion.

[0096] Stability assessment of motor controllers based on electrical shock erosion data: Key performance parameters of the motor controller are extracted from the three-dimensional operating model. A stability assessment index system for the controller is established by combining the erosion impact in the electrical shock erosion data, including voltage ripple coefficient, switching loss rate, and temperature margin. A critical stability threshold is set. The remaining time for the index to reach the critical threshold is calculated by real-time monitoring of the current value and trend of the assessment index. The state with a remaining time of less than 24 hours is defined as the critical support loss state. The corresponding index, remaining time, and erosion causes are recorded to form critical support loss data.

[0097] Risk difference coefficient extrapolation is performed on critical support loss data: Based on the functional importance of each component of the motor drive control system and combined with the remaining time in the critical support loss data, the calculation rule for risk difference coefficient is set: Risk difference coefficient = (1 / remaining time) × component importance weight; For multiple critical indicators of the same component, the largest risk difference coefficient is taken as the final coefficient of that component; By comparing the coefficients of different components, the risk difference of each component is quantified, risk difference coefficient data is formed, and the coefficient ranking is marked.

[0098] Failure probability estimation is performed by combining electrical shock erosion degree data, critical support loss data, and risk difference coefficient: A failure probability assessment model is constructed using the analytic hierarchy process (AHP), with three data points serving as assessment criteria layers: electrical shock erosion degree, remaining time of critical support loss, and risk difference coefficient; scoring standards are set for each criterion layer; scores are assigned based on the actual data of each component, and a weighted total score is calculated; the total score is mapped to the failure probability; the model is validated through historical failure cases, and the final failure probability data is obtained after adjusting the weights.

[0099] It should be noted that step S24 includes the following steps:

[0100] Step S241: Partition the fault probability data to obtain fault risk partition data;

[0101] Step S242: Calculate the current fluctuation difference between different risk areas based on the fault risk zoning data to obtain current fluctuation difference data;

[0102] Step S243: Calculate the temperature rise variance between different risk areas based on the risk zoning data to obtain temperature rise variance data;

[0103] Step S244: Based on the current fluctuation difference data and temperature rise variance data, perform load change energy simulation on the electromechanical operation mechanism data to obtain load change energy data;

[0104] Step S245: Calculate the abnormal impact energy increment based on the load change energy data to obtain abnormal impact energy data.

[0105] Specifically, the fault probability data is partitioned: based on the structure and function of the electric vehicle motor drive control system, the system is divided into four local areas—controller area (including IGBT modules, capacitors, and drive chips), stator area (including stator windings and iron core), rotor shaft area (including rotor, shaft, and bearings), and cooling system area (including cooling pumps, pipes, and radiators). The fault probability of components in each area is extracted from the fault probability data; the maximum value method is used to determine the regional fault probability of each area; based on the regional fault probability, fault risk zones are divided: high-risk zone (probability > 70%), medium-risk zone (30%-70%), and low-risk zone (< 30%). The boundaries and included components of each zone are marked to form fault risk zone data, and the zones are marked with different colors in the three-dimensional operating model.

[0106] Calculate the current fluctuation difference between different risk areas based on fault risk zoning data: Extract current data for each area from the real-time data interface of the 3D operating model—IGBT output current for the controller area, winding input current for the stator area, excitation current for the rotor shaft area, and no current data for the cooling system area; calculate the current fluctuation difference between adjacent areas: fluctuation difference = |current fluctuation value of area A - current fluctuation value of area B|; calculate the fluctuation difference across areas; record the fluctuation difference values, associated areas, and calculation time for all areas to form current fluctuation difference data, and mark the direction of the fluctuation difference with arrows in the 3D model;

[0107] Calculate the temperature rise variance between different risk zones based on fault risk zoning data: Extract temperature data for each zone from the three-dimensional operating model—IGBT temperature for the controller zone, winding temperature for the stator zone, bearing temperature for the rotor shaft zone, and outlet temperature for the cooling system zone; calculate the average temperature rise of each zone over 5 minutes, such as 20℃ for the controller zone, 15℃ for the stator zone, 10℃ for the rotor shaft zone, and 5℃ for the cooling system zone; calculate the temperature rise variance between different zones using the variance calculation formula; calculate the variance by grouping adjacent zones, record the variance values, associated zones, and temperature acquisition time to form temperature rise variance data.

[0108] Load change energy simulation is performed on the operating mechanism data of electrical machinery based on current fluctuation difference data and temperature rise variance data: The correlation model of "current fluctuation - energy loss" and "temperature rise - energy conversion" is extracted from the operating mechanism data of electrical machinery—for example, for every 10A increase in current fluctuation difference, the motor copper loss increases by 50W; for every 10°C increase in temperature rise variance, the motor iron loss increases by 30W. The current fluctuation difference data and temperature rise variance data are substituted into the model to calculate the energy loss under the current load. Load change scenarios are set, and the energy loss under different load rates is calculated based on the correlation between "load rate - current fluctuation - temperature rise" in the mechanism data. Load energy change curves are constructed using energy simulation software (such as PSCAD) (the horizontal axis is the load rate, and the vertical axis is the energy loss), marking the energy change trend under different load rates (e.g., when the load rate > 80%, the energy loss rate increases faster), thus forming load change energy data.

[0109] Abnormal impact energy increment is calculated based on load change energy data: the abnormal impact energy increment is defined as "energy loss at a certain load rate - energy loss at a baseline load rate (50%)"; for example, the energy loss at a load rate of 60% is 149.75W, the baseline loss is 118.75W, and the increment is 31W; at a load rate of 80%, if the energy loss is 200W, the increment is 81.25W; at a load rate of 100%, the energy loss is 280W, and the increment is 161.25W; the impact energy increment at different load rates is calculated for each fault risk zone; the correlation between the increment and the load rate is analyzed, and the increment threshold for high-risk areas is marked based on the fault probability of the area; finally, abnormal impact energy data is generated.

[0110] It should be noted that step S244 includes the following steps:

[0111] Historical operating condition data is acquired, and the historical operating condition data is analyzed for spatiotemporal distribution to obtain load intensity distribution data;

[0112] Stability difference analysis was performed on the operating mechanism data of electrical machinery based on current fluctuation difference data and temperature rise variance data to obtain stability difference data.

[0113] The anti-interference capability is analyzed by the stability difference data to obtain anti-interference capability data;

[0114] Based on the load intensity distribution data, a multi-dimensional impact simulation evaluation is performed on the stability difference data and anti-interference capability data to obtain impact carrying density data.

[0115] Incremental analysis was performed on the impact-carrying density data to obtain incremental impact-carrying density data;

[0116] Based on the current fluctuation difference data and temperature rise variance data, load change energy simulation is performed on the impact-carrying density increment data to obtain load change energy data.

[0117] Specifically, historical operating condition data of the electric vehicle motor drive control system is obtained: operating data from the past 12 months is exported through the on-board T-BOX terminal, covering different seasons (low temperature in winter, high temperature in summer), different road conditions (urban congestion, highways, rural roads), and different driving habits (rapid acceleration, smooth driving, and emergency braking), with a data volume of over 100,000 sets; each data entry includes parameters such as operating condition type, driving time, ambient temperature, motor load rate, current, voltage, and temperature; the historical data is cleaned to remove invalid data caused by equipment failure (such as blank data from offline sensors), retaining over 80,000 sets of valid data to form a historical operating condition dataset.

[0118] Secondly, the historical operating data was analyzed for its spatiotemporal distribution: From a time perspective, the data was divided into periods by month, weekday, and hour, and the operating condition distribution for each period was statistically analyzed (e.g., during winter (January-March), the average motor load rate was 40%; during summer (July-September), the average load rate was 55%; during weekday morning rush hour (7-9 AM), the load rate fluctuated greatly, averaging 50%; and during nighttime (10-6 PM), the load rate was low, averaging 20%). From a spatial perspective, the data was divided by driving area (city center, suburbs, highways), and the operating characteristics of different areas were statistically analyzed (e.g., congestion in city centers, etc.). Load rate fluctuation frequency: 15 times / hour; high-speed uniform speed, fluctuation frequency: 3 times / hour; using a heat map visualization tool, a three-dimensional heat map of "time-space-load rate" is drawn (horizontal axis is time, vertical axis is space, color depth represents load rate), which intuitively shows the spatiotemporal distribution pattern of load intensity (e.g., the urban center area during the morning rush hour on summer weekdays is a red high load area); through statistical analysis, the mean, standard deviation, and peak load rate of each spatiotemporal interval are calculated to form load intensity distribution data that includes time intervals, spatial regions, and load rate statistics.

[0119] Subsequently, the stability differences in the electromechanical operating mechanism data were analyzed based on the current fluctuation difference data and temperature rise variance data: stability parameters (such as the voltage fluctuation tolerance value of the controller capacitor and the temperature stability coefficient of the stator winding) of each region (controller region, stator region, etc.) were extracted from the electromechanical operating mechanism data; the current fluctuation difference data (such as 20A between the controller region and the stator region) and the temperature rise variance data (such as 6.25℃²) were substituted into the stability assessment model to calculate the stability score of each region under the current fluctuation and variance—for example, the voltage fluctuation tolerance value of the controller region is 30A, the current fluctuation difference is 20A, and the score is 80 points; the temperature stability coefficient of the stator region is 10℃², the current variance is 6.25℃², and the score is 75 points; the stability scores of different regions were compared (such as 80 points in the controller region vs. 85 points in the rotor region) to identify regions with weak stability (such as the stator region); the causes of stability differences were analyzed (such as the low score in the stator region being due to winding insulation erosion), and the region name, stability score, and cause of difference were recorded to form stability difference data.

[0120] Next, an anti-interference capability analysis was performed on the stability difference data: Interference test scenarios were set, including power supply voltage fluctuation interference (±10% of rated voltage), electromagnetic radiation interference (100-300MHz frequency band), and temperature interference (ambient temperature ±20℃). The stability difference data for each region (e.g., 75 points for the stator region) were input into the anti-interference test model to simulate the changes in stability scores under different interference scenarios—for example, under power supply voltage fluctuation interference, the controller region score dropped from 80 points to 72 points, and the stator region score dropped from 75 points to 65 points; electromagnetic radiation... Under interference, the controller area score drops to 75 points, and the stator area score drops to 70 points; under temperature interference, the controller area score drops to 78 points, and the stator area score drops to 68 points; calculate the anti-interference capability index for each area (the reciprocal of the difference in score before and after interference, the smaller the difference, the larger the index), for example, the difference in power supply interference in the controller area is 8, and the index is 0.125; the difference in the stator area is 10, and the index is 0.1; by comparing the indices, identify areas with weak anti-interference capabilities (such as the stator area); generate anti-interference capability data including area name, interference type, anti-interference capability index, and weak points.

[0121] Then, based on the load intensity distribution data, a multi-dimensional impact simulation evaluation was conducted on the stability difference data and anti-interference capability data: A multi-dimensional impact evaluation model was constructed, taking load intensity (time-space distribution), stability difference (score), and anti-interference capability (index) as input dimensions; an evaluation scenario was set, such as "summer weekday morning rush hour (high load intensity, load rate 55%) + city center (congestion, high fluctuation) + power supply interference (±10%)"; the load intensity distribution data (load rate 55%, fluctuation frequency 15 times / hour), stability difference data (controller area 80 points, stator area 75 points), and anti-interference capability data (controller area 75 points) under this scenario were analyzed. Input the model with values ​​of 0.125 for the controller region and 0.1 for the stator region. Use the Monte Carlo simulation algorithm to simulate the impact process in this scenario 1000 times and calculate the impact intensity (impact value considering load, stability, and anti-interference) of each region in each simulation. Statistically analyze the distribution of impact intensity (e.g., the impact intensity in the controller region is concentrated in the range of 80-100, and in the stator region it is concentrated in the range of 70-90). Use kernel density estimation to calculate the impact carrying density (the probability of impact intensity distribution per unit time). For example, the peak impact carrying density in the controller region is 0.9 (corresponding to an impact intensity of 90), and the peak value in the stator region is 0.7 (corresponding to an impact intensity of 80), thus forming the impact carrying density data.

[0122] Next, incremental analysis was performed on the impact carryover density data: a baseline scenario (e.g., off-peak hours on winter weekdays, low load intensity 40%, no interference) was selected, and the impact carryover density under this scenario was calculated (e.g., 0.5 in the controller area, 0.4 in the stator area); the density difference between the target scenario (e.g., high load + interference during the early morning rush hour in summer) and the baseline scenario was calculated, i.e., the impact carryover density increment (0.9-0.5=0.4 in the controller area, 0.7-0.4=0.3 in the stator area); the correlation between the increment and scenario parameters was analyzed (e.g., for every 10% increase in load rate, the increment in the controller area increases by 0.1; when there is power supply interference, the increment increases by 0.15); high increment scenarios (e.g., load rate > 60% + power supply interference, increment in the controller area > 0.5) were marked to form the impact carryover density increment data.

[0123] Finally, load change energy simulation was performed on the incremental data of impact-carried density based on the current fluctuation difference data and temperature rise variance data: the current fluctuation difference data and temperature rise variance data (6.25℃²) were used as weighting factors to perform weighted correction on the incremental data of impact-carried density; the corrected incremental data were substituted into the load change energy model, and the energy value under different load rates was calculated by combining the basic energy loss values ​​under different load rates; the "load rate-energy value" curve was generated by the energy simulation software to indicate whether the energy under each load rate exceeded the safety threshold; finally, the load change energy data was formed.

[0124] It should be noted that step S3 includes the following steps:

[0125] Step S31: Normalize the abnormal impact energy data to obtain normalized abnormal impact energy data;

[0126] Step S32: Perform convolution calculation on the normalized abnormal impact energy data to obtain abnormal impact energy convolution data;

[0127] Step S33: Assign a local risk level to the electric vehicle motor drive control system based on the abnormal impact energy convolution data and the fault probability data to obtain a local risk level map.

[0128] It should be noted that step S33 includes the following steps:

[0129] Step S331: Recalibrate the spatial coordinates of the three-dimensional running model to obtain the recalibrated spatial coordinates;

[0130] Step S332: Perform risk linkage analysis on the spatial recalibration coordinates based on the abnormal impact energy convolution data and fault probability data to obtain risk linkage data;

[0131] Step S333: Calculate the multivariate local risk index based on the abnormal impact energy convolution data, fault probability data, and risk linkage data to obtain the multivariate local risk index;

[0132] Step S334: Assign local risk level values ​​to the electric vehicle motor drive control system based on the multivariate local risk index to obtain a local risk level map.

[0133] Specifically, in step S331, the spatial coordinates of the three-dimensional running model are recalibrated: Due to potential component displacement and sensor installation loosening during long-term operation of the motor drive system, the spatial coordinates of the original three-dimensional model may deviate from the actual physical positions. A laser rangefinder is used to measure the key reference points of the motor drive system—six reference points are selected, including the corner of the controller housing, the center of the stator core end face, and the shaft center, and the actual three-dimensional coordinates (X, Y, Z) of each reference point are recorded. The actual coordinates are compared with the reference point coordinates in the original model, and the coordinate deviation value is calculated (e.g., the controller corner coordinates in the original model are (100, 200, 300), but the actual coordinates are (100, 200, 300). The coordinates are (102, 201, 300), with a deviation of (2, 1, 0). A spatial coordinate transformation algorithm (such as Affine transformation) is used to correct the coordinates of the entire 3D model based on the deviation values ​​of the 6 reference points. For example, the X coordinate of all points in the model is increased by 2mm, the Y coordinate is increased by 1mm, and the Z coordinate remains unchanged. After correction, the coordinates of other non-reference points (such as the center of the IGBT module) are verified again using a laser rangefinder to ensure that the deviation is less than 0.5mm. Finally, the spatial recalibrated coordinates that closely match the actual physical position are obtained, and the coordinate system of the 3D model is updated so that each spatial unit in the model can accurately correspond to the physical position of the actual component.

[0134] Risk linkage analysis is performed on the spatial recalibration coordinates based on the convolutional data of abnormal impact energy and the fault probability data: the correlation between each spatial unit is extracted from the spatial recalibration coordinates. For example, unit A in the controller area and unit B in the stator area are connected by wires and have an electrical correlation; unit A and unit C in the cooling system area are in contact with each other through heat sinks and have a thermal correlation. The convolutional data of abnormal impact energy and the fault probability data are input into the correlation analysis model; the risk transmission coefficient between the correlated units is calculated—the risk linkage strength is judged based on the magnitude of the transmission coefficient, such as strong linkage between unit A and B, and strong linkage between unit A and C; all correlated unit pairs, linkage types, transmission coefficients, and linkage strengths are recorded to form risk linkage data, and the correlation relationships are marked with lines of different colors in the 3D model.

[0135] The local risk level of the motor drive system is assigned based on the multivariate local risk index: A risk index threshold is set, dividing the index into 9 risk levels—index ≥ 2.0 is Level 1 (extremely high risk, bright red), 1.8-2.0 is Level 2 (high risk, dark red), 1.6-1.8 is Level 3 (medium-high risk, orange), 1.4-1.6 is Level 4 (medium risk, orange-yellow), 1.2-1.4 is Level 5 (medium-low risk, yellow), 1.0-1.2 is Level 6 (low risk, light yellow), 0.8-1.0 is Level 7 (relatively low risk, light green), 0.6-0.8 is Level 8 (extremely low risk, green), and <0.6 is Level 9 (safe, dark green). The risk index of each spatial unit is compared with the threshold to determine its risk level (e.g., unit A with an index of 2.0 is level 1, unit B with an index of 1.5 is level 4, and unit C with an index of 1.18 is level 6). In the spatially recalibrated 3D model, each unit is assigned a corresponding color according to its level, and a floating information box is added. The levels of all units are integrated to form a local risk level map that covers all local areas of the motor drive system, with accurate coordinates and clear risk levels. In the map, the controller area units with extremely high risk are bright red, the stator area units with medium risk are orange-yellow, and the cooling system area units with low risk are green. The correlation lines (strong linkage) between the red units and the surrounding units are clearly visible, intuitively showing the risk distribution and diffusion relationship.

[0136] Example 2

[0137] This application also discloses a fault diagnosis system for an electric vehicle motor drive control system.

[0138] Reference Figure 2 A fault diagnosis system for an electric vehicle motor drive control system, comprising:

[0139] The data acquisition module is used to acquire real-time operating data through multi-source sensor units deployed in the electric vehicle motor drive control system, and to perform global acquisition of the real-time operating data, obtain a global operating image based on the global acquisition results, and construct a three-dimensional operating model based on the global operating image.

[0140] The analysis and processing module is used to restore the electrical and mechanical coupling relationship of the three-dimensional operation model to obtain operation mechanism data; based on the operation mechanism data and the global operation image, the abnormal probability of the motor drive control system is estimated to obtain fault probability data; and the system load impact energy increment is calculated from the fault probability data to obtain abnormal impact energy data.

[0141] The risk classification module is used to assign local risk levels to the electric vehicle motor drive control system based on the fault probability data and the abnormal impact energy data, and form a local risk level map.

[0142] The optimization deployment module is used to construct a safety hazard identification optimization model for the motor drive control system based on the local risk level map, and send the safety hazard identification optimization model to the on-board diagnostic terminal to perform fault diagnosis of the electric vehicle motor drive control system.

[0143] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention, they should all fall within the protection scope of the present invention.

[0144] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0145] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A fault diagnosis method for an electric vehicle motor drive control system, characterized in that, Includes the following steps: Step S1: Acquire real-time operating data through multi-source sensor units deployed in the electric vehicle motor drive control system, and perform global acquisition of the real-time operating data. Obtain a global operating image based on the global acquisition results, and construct a three-dimensional operating model based on the global operating image. Step S1 includes the following steps: Step S11: By using sensor units deployed on the motor controller, motor stator, and motor shaft, the global operating data of the electric vehicle motor drive control system under different operating conditions is collected in real time to form a global operating image; Step S12: Perform data normalization and distribution harmonic processing on the global running image, and obtain the running state harmonic image based on the processing result; Step S13: Mark the running feature points on the running status harmonization image. The running feature points include motor current fluctuation points, voltage drop points, speed abnormal points, and temperature rise points, to obtain running feature point marking data. Step S14: Based on the running state harmonized image and the running feature point marker data, construct a three-dimensional running model of the electric vehicle motor drive control system, and obtain a three-dimensional running model based on the construction results; Step S2: Reconstruct the electrical and mechanical coupling relationship of the three-dimensional operation model to obtain operation mechanism data; Based on the operation mechanism data and the global operation image, estimate the anomaly probability of the motor drive control system to obtain fault probability data; Calculate the load impact energy increment of the fault probability data to obtain abnormal impact energy data. Step S3: Assign local risk level values ​​to the electric vehicle motor drive control system based on the fault probability data and the abnormal impact energy data to form a local risk level map; Step S4: Construct a safety hazard identification and optimization model for the motor drive control system based on the local risk level map, and send the safety hazard identification and optimization model to the on-board diagnostic terminal to perform fault diagnosis of the electric vehicle motor drive control system.

2. The fault diagnosis method for an electric vehicle motor drive control system according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: Perform anomaly analysis on the motor current and voltage trends of the three-dimensional operating model to obtain electrical characteristic trend data; Step S22: Based on the electrical characteristic directional data, the electrical and mechanical coupling mechanism of the three-dimensional operation model is restored to obtain the electrical and mechanical operation mechanism data; Step S23: Based on the electromechanical operation mechanism data and the global operation image, perform anomaly probability estimation on the electric vehicle motor drive control system to obtain fault probability data; Step S24: Calculate the abnormal impact energy increment on the fault probability data to obtain abnormal impact energy data.

3. The fault diagnosis method for an electric vehicle motor drive control system according to claim 2, characterized in that, Step S23 includes the following steps: Step S231: Perform vehicle power supply stability distribution analysis on the global operation image to obtain the power supply fluctuation distribution density; Step S232: Based on the electromechanical operation mechanism data, decompose the power supply fluctuation distribution density and the three-dimensional operation model to obtain the electrical shock erosion degree data; Step S233: Based on the electrical shock erosion degree data, perform a stability assessment on the motor controller to obtain critical support loss data; Step S234: Perform risk difference coefficient deduction on the critical support loss data to obtain risk difference coefficient data; Step S235: Based on the electrical shock erosion degree data, the critical support loss data, and the risk difference coefficient, the failure probability is estimated to obtain failure probability data.

4. The fault diagnosis method for an electric vehicle motor drive control system according to claim 2, characterized in that, Step S24 includes the following steps: Step S241: Partition the fault probability data to obtain fault risk partition data; Step S242: Calculate the current fluctuation difference between different risk areas based on the fault risk zoning data to obtain current fluctuation difference data; Step S243: Calculate the temperature rise variance between different risk areas based on the risk zoning data to obtain temperature rise variance data; Step S244: Based on the current fluctuation difference data and temperature rise variance data, perform load change energy simulation on the electromechanical operation mechanism data to obtain load change energy data; Step S245: Calculate the abnormal impact energy increment based on the load change energy data to obtain abnormal impact energy data.

5. The fault diagnosis method for an electric vehicle motor drive control system according to claim 4, characterized in that, Step S244 includes the following steps: Historical operating condition data is acquired, and the historical operating condition data is analyzed for spatiotemporal distribution to obtain load intensity distribution data; Stability difference analysis was performed on the operating mechanism data of electrical machinery based on current fluctuation difference data and temperature rise variance data to obtain stability difference data. The anti-interference capability is analyzed by the stability difference data to obtain anti-interference capability data; Based on the load intensity distribution data, a multi-dimensional impact simulation evaluation is performed on the stability difference data and anti-interference capability data to obtain impact carrying density data. Incremental analysis was performed on the impact-carrying density data to obtain incremental impact-carrying density data; Based on the current fluctuation difference data and temperature rise variance data, load change energy simulation is performed on the impact-carrying density increment data to obtain load change energy data.

6. The fault diagnosis method for an electric vehicle motor drive control system according to claim 1, characterized in that, Step S3 includes the following steps: Step S31: Normalize the abnormal impact energy data to obtain normalized abnormal impact energy data; Step S32: Perform convolution calculation on the normalized abnormal impact energy data to obtain abnormal impact energy convolution data; Step S33: Assign a local risk level to the electric vehicle motor drive control system based on the abnormal impact energy convolution data and the fault probability data to obtain a local risk level map.

7. The fault diagnosis method for an electric vehicle motor drive control system according to claim 6, characterized in that, Step S33 includes the following steps: Step S331: Recalibrate the spatial coordinates of the three-dimensional running model to obtain the recalibrated spatial coordinates; Step S332: Perform risk linkage analysis on the spatial recalibration coordinates based on the abnormal impact energy convolution data and fault probability data to obtain risk linkage data; Step S333: Calculate the multivariate local risk index based on the abnormal impact energy convolution data, fault probability data, and risk linkage data to obtain the multivariate local risk index; Step S334: Assign local risk level values ​​to the electric vehicle motor drive control system based on the multivariate local risk index to obtain a local risk level map.

8. A fault diagnosis system for an electric vehicle motor drive control system, applied to the fault diagnosis method for an electric vehicle motor drive control system according to any one of claims 1-7, characterized in that, include: The data acquisition module is used to acquire real-time operating data through multi-source sensor units deployed in the electric vehicle motor drive control system, and to perform global acquisition of the real-time operating data, obtain a global operating image based on the global acquisition results, and construct a three-dimensional operating model based on the global operating image. The analysis and processing module is used to reconstruct the electrical and mechanical coupling relationship of the three-dimensional operating model to obtain operating mechanism data; Based on the aforementioned operating mechanism data and global operating image, anomaly probability estimation is performed on the motor drive control system to obtain fault probability data. Then, the system load impact energy increment is calculated on the fault probability data to obtain abnormal impact energy data. The risk classification module is used to assign local risk levels to the electric vehicle motor drive control system based on the fault probability data and the abnormal impact energy data, and form a local risk level map. The optimization deployment module is used to construct a safety hazard identification optimization model for the motor drive control system based on the local risk level map, and send the safety hazard identification optimization model to the on-board diagnostic terminal to perform fault diagnosis of the electric vehicle motor drive control system.

9. A computer-readable storage medium, characterized in that: The system stores instructions that, when executed on a computer, cause the computer to perform a fault diagnosis method for an electric vehicle motor drive control system as described in any one of claims 1 to 7.

Images