Method and system for testing motor driver of vehicle-mounted motor of new energy automobile

Abstract

The invention discloses a test method and system for a motor driver of a vehicle-mounted motor of a new energy automobile, and the method comprises the steps: S1, collecting dynamic environment and working condition parameters: collecting the performance parameters of the motor driver under a preset dynamic environment condition and a simulated working condition; the dynamic environment conditions comprise the combination of the temperature of-40 DEG C to 85 DEG C, the humidity of 10%-95% and the vibration frequency of 10-2000 Hz; the simulation working conditions comprise an NEDC circulation working condition, a WLTC circulation working condition and a rapid acceleration / deceleration dynamic working condition; s2, multi-modal anomaly detection: on the basis of the parameters acquired in the step S1, through a multi-modal anomaly detection algorithm fusing a random forest, an isolated forest and an auto-encoder, and in combination with a Bayesian optimization dynamic threshold, identifying an abnormal mode of the motor driver; s3, performing adaptive parameter optimization and performance evaluation; and S4, generating a test result. The method has the advantages that the technical scheme covers the temperature of-40 DEG C to 85 DEG C, the vibration of 10 Hz to 2000 Hz and the dynamic working condition, the reliability verification scene is expanded to the full life cycle from the static state, and the defect that the environment and working condition coverage is incomplete in the original scheme is overcome.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automobile motor control, and particularly relates to a test method and system for a motor driver of a vehicle-mounted motor of a new energy automobile. BACKGROUND

[0002] As a core component of the power system, the performance of the new energy automobile motor driver directly determines the power performance, safety and reliability of the whole vehicle. With the development of high-voltage platforms (800V) and high-power density (>3kW / kg) technologies, the driver needs to operate stably for a long time under extreme environments (-40℃-85℃ temperature, 10%-95% humidity, 10-2000Hz vibration) and complex working conditions (sudden acceleration, climbing, frequent start-stop), and the traditional test method has been difficult to meet the needs.

[0003] The existing technology has three limitations: first, the fault simulation is one-sided, and it focuses on single electrical faults (such as overcurrent), and lacks injection verification of complex scenarios such as communication faults (CAN bus errors) and mechanical faults (bearing jam); second, the life cycle coverage is insufficient, and it focuses on performance testing in the research and development stage, and lacks full-cycle verification of production consistency, 5000h aging attenuation and retirement residual value evaluation; third, the dynamic adaptability is weak, and the environmental stress and working condition load are statically set, and dynamic parameter optimization under the coupling of "fault-environment-load" is not realized.

[0004] Therefore, there is an urgent need for a test method that integrates fault injection, full life cycle verification and dynamic monitoring to comprehensively evaluate the reliability and durability of the driver under complex scenarios and support the improvement of the safety performance of new energy automobiles. SUMMARY

[0005] The technical problem solved by the present application is to provide a test method and system for a motor driver of a vehicle-mounted motor of a new energy automobile, which solves one or more of the above-mentioned problems in the prior art.

[0006] To solve the above technical problems, one technical solution adopted by the present application is: a test method for a motor driver of a vehicle-mounted motor of a new energy automobile, the innovation point of which is that it comprises the following steps:

[0007] S1. Dynamic environment and working condition parameter acquisition: under the preset dynamic environment conditions and simulated working conditions, the performance parameters of the motor driver are collected; the dynamic environment conditions include combinations of temperature -40℃-85℃, humidity 10%-95% and vibration frequency 10-2000Hz; the simulated working conditions include NEDC cycle working conditions, WLTC cycle working conditions and sudden acceleration / deceleration dynamic working conditions;

[0008] S2. Multimodal anomaly detection: Based on the parameters collected in step S1, a multimodal anomaly detection algorithm is used to identify abnormal patterns in the motor driver. This algorithm combines random forest, isolation forest, and autoencoder, and uses Bayesian optimization to dynamically adjust the threshold value.

[0009] S3. Adaptive parameter optimization and performance evaluation: An adaptive termination mechanism is used to optimize the control parameters of the motor driver. This mechanism includes a convergence threshold condition (the change rate of fitness score is less than 0.5% for 5 consecutive iterations) and a resource threshold condition (parameter partition search is triggered when the number of iterations reaches the upper limit). The optimization process is accelerated by a GPU parallel computing framework.

[0010] S4. Test result generation: Based on the anomaly detection results and the performance data after parameter optimization, a reliability evaluation report of the motor driver is generated.

[0011] In some embodiments, the specific implementation of the multimodal anomaly detection algorithm in step S2 includes:

[0012] Random forest algorithm is used to process the trend anomalies of regular performance parameters;

[0013] Isolation forest algorithm is used to detect isolated point anomalies in high-dimensional data;

[0014] Autoencoder is used to identify unstructured noise interference anomalies through reconstruction error;

[0015] Bayesian optimization is used to dynamically adjust the anomaly judgment threshold value of each algorithm. The threshold adjustment period is positively correlated with the change rate of environmental temperature.

[0016] In some embodiments, the dynamic environmental conditions in step S1 are achieved by integrating high and low temperature and humidity chambers and vibration tables. The simulation conditions are injected into the control signals through the linkage of the simulation model built by MATLAB / Simulink and the dSPACE real-time simulator.

[0017] In some embodiments, the GPU parallel computing framework in step S3 supports parallel processing of k-fold cross-validation, and simultaneously verifies at least 3 groups of parameter combinations.

[0018] A motor driver test system for a new energy vehicle on-board motor, comprising:

[0019] Environmental simulation module: including high and low temperature and humidity chamber (temperature control range -40℃~85℃), vibration table (vibration frequency 10-2000Hz) and humidity sensor, used for simulating dynamic environmental conditions;

[0020] Working condition simulation module: including MATLAB / Simulink working condition model library, dSPACE real-time simulator and vehicle controller interface, used for generating NEDC, WLTC and dynamic working condition control signals;

[0021] Multi-modal anomaly detection module: integrating random forest processing unit, isolated forest processing unit, auto-encoder processing unit and Bayesian threshold optimization unit, used for performing step S2 anomaly detection;

[0022] Parameter optimization module: including adaptive termination control unit (built-in convergence threshold and resource threshold logic) and GPU parallel computing unit, used for performing step S3 parameter optimization;

[0023] Data acquisition and analysis module: connecting voltage, current and temperature sensors of the motor driver, real-time acquisition of performance parameters and generation of test report.

[0024] In some embodiments, the auto-encoder processing unit of the multi-modal anomaly detection module adopts a 3-layer encoder-decoder structure, and the input layer dimension is 12-dimensional key parameters of the motor driver (including bus voltage, phase current, rotor speed and winding temperature).

[0025] In some embodiments, the adaptive termination control unit of the parameter optimization module further includes a real-time performance feedback subunit for dynamically adjusting the iteration depth according to the current fitness score change trend.

[0026] In some embodiments, the environment simulation module and the working condition simulation module realize data synchronization through industrial Ethernet, and the synchronization delay is ≤10ms.

[0027] The beneficial effects of the present application are: test comprehensiveness is improved: covering -40℃ ~ 85℃ temperature, 10-2000Hz vibration and dynamic working condition, reliability verification scene is expanded from static to full life cycle, solving the defects of the original scheme "environment and working condition coverage is not complete".

[0028] Anomaly detection robustness is enhanced: multi-modal algorithm fusion makes the anomaly detection accuracy rate increase by 15% ~ 20%, the false detection rate is reduced to below 5%, especially the identification ability of transient failure and noise interference under extreme environment is significantly improved.

[0029] Test efficiency and precision balance: adaptive optimization mechanism + GPU parallel computing makes the parameter optimization time shorten by 40% ~ 60%, at the same time the fitness score is improved to above 0.9, avoiding the resource waste or insufficient precision caused by traditional fixed iteration number.

[0030] High engineering application value: It can be directly integrated into the factory testing stage of the new energy vehicle motor drive production line, or used for after-sales fault diagnosis. It is expected to reduce the actual vehicle failure rate by 12% to 18% and shorten the product development cycle by more than 30%. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in 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, wherein:

[0032] Figure 1 This is a flowchart of a test method for a motor driver of an on-board motor in a new energy vehicle according to the present invention. Detailed Implementation

[0033] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] This invention includes a testing method and system for motor drivers in new energy vehicles. Through the collaborative operation of three core modules—dynamic environment simulation, multimodal anomaly detection, and adaptive parameter optimization—it achieves accurate evaluation of the motor driver's performance throughout its entire lifecycle. The implementation of the technical solution is described in detail below, combining specific hardware configurations, algorithm flows, and control logic.

[0035] 1. System Components

[0036] The core components of the environmental simulation module include a high and low temperature humidity chamber, an electromagnetic vibration table, and temperature and humidity sensors. The high and low temperature humidity chamber has a temperature control range of -40℃ to 85℃ and a humidity adjustment range of 10% to 95%, and can achieve dynamic control of the temperature cycling rate at 5℃ / min. The electromagnetic vibration table supports a frequency range of 10-2000Hz and a maximum acceleration of 50g, meeting the vibration testing requirements of the ISO16750 standard. The temperature and humidity sensors have a measurement accuracy of ±0.5℃ (temperature) and ±2%RH (humidity), acquiring environmental parameters in real time and feeding them back to the control system to ensure that the simulated environment is consistent with the preset conditions. The main function of this module is to simulate the environmental conditions of new energy vehicles in extreme scenarios such as high altitudes, extreme cold, and coastal areas, and to verify the stability of the motor driver under the coupled effects of temperature, humidity, and vibration.

[0037] The working condition simulation module is composed of a MATLAB / Simulink working condition model library, a dSPACE real-time simulator, and a vehicle controller interface. The model library contains NEDC (New European Driving Cycle), WLTC (Worldwide Light Vehicle Test Cycle), CLTC-P (China Light Duty Vehicle Driving Cycle), and other standard cycle conditions, and supports parameter configuration of custom dynamic conditions such as sudden acceleration, sudden deceleration, and hill driving. The dSPACE real-time simulator (Model MicroAutoBox III) uses FPGA chips to achieve microsecond-level signal processing and can inject real vehicle control signals (such as accelerator pedal opening, brake signal, and shift command) into the motor driver. The vehicle controller interface uses the CAN FD communication protocol with a maximum data transmission rate of 8 Mbps to ensure real-time interaction between working condition signals and motor driver states. The function of this module is to reproduce the dynamic load changes in actual vehicle driving and evaluate the response characteristics of the driver under complex conditions.

[0038] The hardware basis of the multi-modal anomaly detection module is an industrial control computer (configured with NVIDIA A100 GPU and Intel i9-13900K CPU), and the software algorithm is developed based on the Python+TensorFlow framework. The module integrates three parallel processing units: the random forest processing unit is responsible for analyzing the trend anomalies of current, voltage, and other regular parameters (such as slow drift of current caused by continuous overload); the isolation forest processing unit detects isolated points in high-dimensional data (such as transient overvoltage pulses) by constructing random hyperplanes; the autoencoder processing unit uses a 3-layer encoder-decoder structure (input layer 12-dimensional parameters → hidden layer 8-dimensional features → output layer 4-dimensional reconstructed signals) to identify unstructured disturbances such as sensor noise through reconstruction error; in addition, the module also includes a Bayesian threshold optimization unit that can dynamically adjust the anomaly judgment threshold of each algorithm according to the environmental temperature change rate (such as temperature drift from -40°C to 85°C), and the threshold update period is positively related to the temperature change rate (maximum update frequency 1 Hz).

[0039] The parameter optimization module includes an adaptive termination control unit and a GPU parallel computing unit. The adaptive termination control unit is based on FPGA (Xilinx Zynq UltraScale+) and has built-in convergence threshold and resource threshold double logic: when the change rate of fitness score of consecutive 5 iterations is <0.5%, convergence termination is triggered; when the number of iterations reaches 50 and still does not converge, parameter partition search is automatically started (voltage vector angle, carrier frequency, etc. are divided into 3 sub-intervals for independent optimization); the GPU parallel computing unit uses CUDA 12.0 architecture and supports k-fold cross-validation parallel processing, which can simultaneously verify the performance of 8 groups of parameter combinations (such as voltage vector angle ±5°, carrier frequency 5-15 kHz), greatly shortening the optimization time.

[0040] The core device of the data acquisition module is a high-speed data acquisition card (NIPXIe-6368) with a sampling rate of 1 MS / s and 16-bit precision, which can synchronously acquire 12-dimensional key parameters of the motor driver, including bus voltage (range 0-500 V, precision 0.1% FS), phase current (range 0-800 A, precision 0.1% FS), winding temperature (range -50℃-150℃, precision ±1℃), output torque (range 0-500 N·m, precision ±0.5% FS), etc. The acquired data is transmitted in real time to a storage server (capacity 10 TB, read / write speed 1 GB / s) through a PCIe bus, providing raw data support for subsequent anomaly detection and parameter optimization.

[0041] 2. Test method (such as Figure 1 )

[0042] Step S1: Dynamic environment and working condition parameter acquisition

[0043] Environment simulation process: First, the control system presets the environmental parameters, for example, the low-temperature test process is "-40℃ for 2h→ warmed up to 25℃ at a rate of 5℃ / min→ kept for 1h", and at the same time, the electromagnetic vibration table is started to apply a random vibration of 1000Hz, 20g acceleration, and the humidity is maintained at 95% to simulate the coastal high-humidity environment; the temperature and humidity sensor monitors the environment in the box in real time, and when the actual value deviates from the preset value by more than ±1℃ or ±5%RH, the control system adjusts the heating pipe, humidifier and vibration table amplitude through the PID algorithm to ensure the stability of the environmental parameters.

[0044] Working condition simulation process: At the same time of environment simulation, the working condition simulation module sends control instructions to the motor driver through the dSPACE real-time simulator; taking the NEDC cycle working condition as an example, the simulator generates motor speed command (0-12000rpm) and torque demand (0-200N·m) according to the speed-time curve, which is transmitted to the control board card of the driver through the CAN FD bus; the data acquisition module synchronously records the input and output parameters of the driver, with a sampling interval of 1ms, ensuring the capture of transient dynamic process (such as current peak value during sudden acceleration).

[0045] Step S2: Multi-modal anomaly detection

[0046] Algorithm fusion logic: the random forest processing unit first extracts features from the collected 12-dimensional parameters, calculates the probability distribution of the normal state through 100 decision trees (each tree has a depth of no more than 15 layers), and marks a trend anomaly when the probability value of a parameter is less than 0.05; the isolated forest processing unit constructs 100 random hyperplanes for high-dimensional data (such as voltage, current, and torque instantaneous fluctuations), and determines an isolated point anomaly if a sample is isolated with a path length less than 50% of the average length; the autoencoder processing unit identifies noise interference through reconstruction error, and triggers an unstructured anomaly alarm when the mean square error (MSE) between the input signal and the reconstructed signal exceeds 0.02; the Bayesian threshold optimization unit dynamically adjusts the weights of the three algorithms according to the environmental temperature change rate, for example, in a low temperature environment of -40°C, the weight of the isolated forest is increased from 0.3 to 0.4 to enhance the sensitivity to transient pulse interference.

[0047] Step S3: adaptive parameter optimization

[0048] Optimization target and fitness function: taking the efficiency (η) and response time (t_response) of the motor driver as the core optimization target, a fitness function is constructed:

[0049] Fitness score = 0.6 x η + 0.4 x (1 - t response / t max )

[0050] Where t max is the preset maximum response time (200 ms), and the efficiency η is calculated by the ratio of input power (voltage x current) to output power (torque x speed / 9550).

[0051] Adaptive termination and parallel acceleration: after the parameter optimization is started, the adaptive termination control unit monitors the fitness score changes in real time, and if the change rate of the last 5 iterations is <0.5% (such as from 0.91 to 0.914), it is determined that the parameters have converged and the optimization is terminated; if the score does not converge after 50 iterations (such as fluctuating between 0.88 and 0.90), the parameter partition search is triggered, and parameters such as voltage vector angle (-15°-15°) and carrier frequency (5-15 kHz) are divided into 3 subintervals, and each subinterval independently performs optimization iteration; the GPU parallel computing unit realizes the synchronous verification of 8 groups of parameter combinations through CUDA kernel functions, and the single optimization time is shortened from 2 hours of traditional serial calculation to 40 minutes.

[0052] Examples and comparative examples

[0053] 1. Example 1: comprehensive performance test under basic configuration

[0054] Test object: a certain brand 200kW permanent magnet synchronous motor driver (rated voltage 350V, maximum current 600A);

[0055] Environmental conditions: normal temperature (25℃), humidity 50%, no vibration;

[0056] Working condition setting: NEDC cycle working condition + rated load (200N·m);

[0057] Algorithm configuration: multi-modal anomaly detection (random forest + isolated forest + self-encoder) + adaptive parameter optimization (convergence threshold termination).

[0058] 2. Example 2: Reliability test under extreme environment

[0059] Test object: the motor driver of example 1;

[0060] Environmental conditions: low temperature (-40℃, 2h insulation) + vibration (1000Hz, acceleration 30g) + humidity 95%;

[0061] Working condition setting: WLTC cycle working condition + overload load (250N·m, lasting 10min);

[0062] Algorithm configuration: multi-modal anomaly detection (Bayesian threshold dynamic adjustment) + adaptive parameter optimization (resource threshold triggering partition search).

[0063] 3. Example 3: Parameter optimization efficiency test under dynamic working condition

[0064] Test object: the motor driver of example 1;

[0065] Environmental conditions: high temperature (85℃) + humidity 30%;

[0066] Working condition setting: sudden acceleration / deceleration working condition (0-100km / h acceleration→braking, repeated 20 times);

[0067] Algorithm configuration: multi-modal anomaly detection + GPU parallel optimization (8 groups of parameters verified at the same time).

[0068] 4. Comparative example (prior art scheme)

[0069] Test object: the motor driver of example 1;

[0070] Environmental conditions: normal temperature (25℃), no vibration, humidity 50% (static environment);

[0071] Working condition setting: rated load (200N·m) static test;

[0072] Algorithm configuration: single random forest anomaly detection (no dynamic threshold) + fixed iteration number (50 times) parameter optimization (no parallel acceleration).

[0073] Performance comparison data table

[0074]

[0075]

[0076] The working principle of the technical solution is:

[0077] Multi-modal anomaly detection: through the algorithm fusion of random forest, isolated forest and self-encoder, respectively covering trend, isolated point and unstructured anomaly mode, combined with Bayesian dynamic threshold optimization, the problem of insufficient adaptability of single algorithm to complex scene is solved.

[0078] Dynamic environment and working condition simulation: linkage of environment simulation module and working condition simulation module, reproduction of coupling effect of environmental factors such as temperature, humidity and vibration and dynamic working conditions such as sudden acceleration and overload, filling the gap of traditional static test in covering actual scene.

[0079] Adaptive parameter optimization: dynamic termination mechanism based on convergence trend and resource threshold, combined with GPU parallel computing, realizing the balance of parameter optimization accuracy and efficiency, avoiding over-optimization or under-optimization caused by fixed iteration number.

[0080] The technical solution has the advantages of:

[0081] Test comprehensiveness is improved: covering -40℃ ~ 85℃ temperature, 10-2000Hz vibration and dynamic working condition, reliability verification scene is expanded from static to full life cycle, solving the defect of "incomplete environment and working condition coverage" of the original scheme.

[0082] Robustness of anomaly detection is enhanced: multi-modal algorithm fusion makes the accuracy of anomaly detection improve by 15% ~ 20%, the false detection rate is reduced to below 5%, especially the identification ability of instantaneous failure and noise interference in extreme environment is significantly improved.

[0083] Test efficiency and precision balance: adaptive optimization mechanism + GPU parallel computing makes the parameter optimization time shorten by 40% ~ 60%, at the same time, the fitness score is improved to above 0.9, avoiding the waste of resources or insufficient precision caused by traditional fixed iteration number.

[0084] High engineering application value: can be directly integrated into the factory detection link of new energy vehicle motor driver production line, or used for after-sales fault diagnosis, which is expected to reduce the real vehicle failure rate by 12% ~ 18% and shorten the product development cycle by more than 30%.

[0085] Those skilled in the art should understand that the above-mentioned embodiments are only the preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0086] The above-mentioned are only the embodiments of the present application, and do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation, or direct or indirect application in other related technical fields made by using the content of the present application shall be included in the patent protection scope of the present application.

Claims

1. A test method for a motor driver of an on-board motor in a new energy vehicle, characterized in that: Comprising the following steps: S1. Dynamic environment and working condition parameter collection: Under the preset dynamic environment conditions and simulated working conditions, the performance parameters of the motor driver are collected; the dynamic environment conditions include the combination of temperature -40℃~85℃, humidity 10%~95%, and vibration frequency 10-2000Hz; the simulated working conditions include NEDC cycle working condition, WLTC cycle working condition, and sudden acceleration / deceleration dynamic working condition; S2. Multi-modal anomaly detection: Based on the parameters collected in step S1, through the multi-modal anomaly detection algorithm of random forest, isolated forest, and self-encoder, combined with the dynamic threshold of Bayesian optimization, the abnormal mode of the motor driver is identified; S3. Adaptive parameter optimization and performance evaluation: The adaptive termination mechanism is used to optimize the control parameters of the motor driver, which includes the convergence threshold condition (the change rate of fitness score of continuous 5 iterations <0.5%) and the resource threshold condition (the number of iterations reaches the upper limit, which triggers the parameter partition search), and the optimization process is accelerated through the GPU parallel computing framework; S4. Test result generation: According to the anomaly detection results and the performance data after parameter optimization, the reliability evaluation report of the motor driver is generated. 2.The test method of the motor driver of the on-board motor of a new energy vehicle according to claim 1, characterized in that: The specific implementation of the multi-modal anomaly detection algorithm in step S2 includes: Using random forest algorithm to process the trend anomaly of regular performance parameters; Using isolated forest algorithm to detect isolated point anomaly in high-dimensional data; Using self-encoder to identify unstructured noise interference anomaly through reconstruction error; Adjusting the anomaly judgment threshold of each algorithm dynamically through Bayesian optimization, and the threshold adjustment period is positively correlated with the change rate of environmental temperature.

3. The test method of the motor driver of the on-board motor of the new energy vehicle according to claim 1, characterized in that: The dynamic environment conditions in step S1 are realized by integrating high and low temperature and humidity chamber and vibration table, and the simulated working conditions are realized by linking control signals through the working condition model built by MATLAB / Simulink and dSPACE real-time simulator.

4. The test method of the motor driver of the on-board motor of the new energy vehicle according to claim 1, characterized in that: The GPU parallel computing framework in step S3 supports parallel processing of k-fold cross-validation, and at least 3 groups of parameter combinations are verified synchronously.

5. A motor drive test system for a new energy vehicle on-board motor, which implements the test method of any one of claims 1-4, characterized in that: Comprising: Environment simulation module: including high and low temperature and humidity chamber (temperature control range -40℃~85℃), vibration table (vibration frequency 10-2000Hz) and humidity sensor, used for simulating dynamic environment conditions; Working condition simulation module: including MATLAB / Simulink working condition model library, dSPACE real-time simulator and vehicle controller interface, used for generating NEDC, WLTC and dynamic working condition control signals; Multi-modal anomaly detection module: integrating random forest processing unit, isolated forest processing unit, self-encoder processing unit and Bayesian threshold optimization unit, used for executing step S2 anomaly detection; Parameter optimization module: including adaptive termination control unit (built-in convergence threshold and resource threshold logic) and GPU parallel computing unit, used for executing step S3 parameter optimization; Data acquisition and analysis module: connecting voltage, current, temperature sensors of the motor driver, real-time collecting performance parameters and generating test report.

6. The motor drive test system of a new energy vehicle on-board motor according to claim 5, characterized in that: The self-encoder processing unit of the multi-modal anomaly detection module adopts a 3-layer encoder-decoder structure, and the input layer dimension is 12-dimensional key parameters of the motor driver, including bus voltage, phase current, rotor speed and winding temperature.

7. The motor drive test system of a new energy vehicle on-board motor according to claim 5, characterized in that: The adaptive termination control unit of the parameter optimization module further comprises a real-time performance feedback subunit for dynamically adjusting the iteration depth according to the current fitness score change trend.

8. The motor drive test system of a new energy vehicle on-board motor according to claim 5, characterized in that: The environment simulation module and the working condition simulation module realize data synchronization through industrial Ethernet, and the synchronization delay is less than or equal to 10 ms.

Images