Method and system for optimizing controller performance of an intelligent vehicle
By constructing a controller state matrix and performing multi-dimensional performance evaluation, the problem of adaptive and online dynamic optimization of vehicle controllers was solved, improving the adaptive adjustment capability and comprehensiveness of performance evaluation of vehicle controllers, and achieving higher control quality and dynamic response accuracy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HANGZHOU INNOVATIVE DRIVE TECH CO LTD
- Filing Date
- 2026-03-27
- Publication Date
- 2026-06-23
AI Technical Summary
Existing technologies lack a comprehensive, multi-dimensional evaluation system for vehicle controllers, and it is difficult to adaptively and dynamically optimize them online based on the real-time operating status of the vehicle and the response to control commands. This leads to the controller performance deviating from expectations, affecting the vehicle's driving safety, comfort, and economy.
By acquiring vehicle operation data and control command data, a controller state matrix is constructed to conduct multi-dimensional performance evaluation, including evaluation of response sensitivity, error and stability. Based on the evaluation results, adaptive optimization and adjustment are performed to generate performance warning signals.
This enhances the vehicle controller's adaptive adjustment capabilities and the comprehensiveness of its multi-dimensional performance evaluation during actual operation, thereby improving control quality and dynamic response accuracy.
Smart Images
Figure CN122260780A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of vehicle control optimization technology, specifically to a method and system for optimizing the performance of controllers in intelligent vehicles. Background Technology
[0002] Intelligent vehicles integrate multiple complex systems, including environmental perception, decision-making, planning, and execution control. The core of these systems lies in the vehicle control system. The vehicle controller, acting as the "brain" of the intelligent vehicle, is responsible for interpreting upper-level decision commands and precisely controlling the vehicle's throttle, braking, steering, and other actuators. Its control performance directly determines the vehicle's driving safety, comfort, and economy. However, traditional vehicle controller parameter calibration and optimization are mostly based on offline maps and fixed operating conditions. In actual vehicle operation, factors such as road environment, vehicle load, tire wear, and system aging all exhibit high uncertainty and time-varying characteristics. Mismatches often arise between fixed control parameters and constantly changing system states, causing the controller's actual performance to deviate from expectations. For example, problems such as dynamic response hysteresis, increased steady-state error, increased energy consumption, and even control instability may occur, severely restricting the overall performance of the intelligent vehicle. Existing controller performance monitoring mostly focuses only on response speed or steady-state accuracy, lacking a comprehensive, multi-dimensional evaluation system for controller performance. Furthermore, it is difficult to adaptively and dynamically optimize the controller online based on the vehicle's real-time operating status and the response to control commands.
[0003] Therefore, current technologies suffer from a lack of comprehensive, multi-dimensional evaluation systems for controller performance and difficulties in adaptively and dynamically optimizing the controller online based on the vehicle's real-time operating status and control command response. Summary of the Invention
[0004] This application provides a method and system for optimizing the controller performance of intelligent vehicles, which solves the technical problems in the prior art of lacking a comprehensive and multi-dimensional evaluation system for controller performance and the difficulty in adaptively and dynamically optimizing the controller online based on the real-time operating status of the vehicle and the response of control commands. It achieves the technical effect of improving the adaptive adjustment capability of the vehicle controller in actual operation, the comprehensiveness of multi-dimensional performance evaluation, and the control quality and dynamic response accuracy throughout the entire life cycle.
[0005] This application provides a method for optimizing the controller performance of an intelligent vehicle. The method includes: acquiring vehicle operation data and vehicle control command data of a target vehicle; constructing a controller state matrix based on the vehicle operation data; performing a multi-dimensional performance evaluation of the controller based on the vehicle control command data and the controller state matrix, and obtaining the controller performance evaluation result; and adaptively optimizing and adjusting the vehicle controller based on the controller performance evaluation result.
[0006] In possible implementations, acquiring vehicle operation data and vehicle control command data of the target vehicle includes: acquiring the target vehicle's operation parameter acquisition results and control command acquisition results; performing data validity verification and outlier removal on the operation parameter acquisition results to generate the vehicle operation data; and performing data validity verification and outlier removal on the control command acquisition results to generate the vehicle control command data.
[0007] In a possible implementation, constructing a controller state matrix based on the vehicle operation data includes: performing time alignment and data normalization on the vehicle operation data to generate a standardized vehicle operation data sequence; and performing matrix mapping on the standardized vehicle operation data sequence to generate the controller state matrix.
[0008] In a possible implementation, a multi-dimensional performance evaluation of the controller is performed based on the vehicle control command data and the controller state matrix to obtain the controller performance evaluation result. This includes: constructing a vehicle control command matrix based on the vehicle control command data; evaluating the control response sensitivity based on the vehicle control command matrix and the controller state matrix to obtain a control response sensitivity coefficient; evaluating the control error based on the vehicle control command matrix and the controller state matrix to obtain a control error coefficient; evaluating the control stability based on the vehicle control command matrix and the controller state matrix to obtain a control stability coefficient; and outputting the control response sensitivity coefficient, the control error coefficient, and the control stability coefficient as the controller performance evaluation result.
[0009] In a possible implementation, evaluating the control response sensitivity based on the vehicle control command matrix and the controller state matrix to obtain the control response sensitivity coefficient includes: analyzing the control response delay based on the vehicle control command matrix and the controller state matrix to obtain the control response delay parameter; training the control response sensitivity evaluation network based on historical control response sensitivity evaluation records; inputting the control response delay parameter into the control response sensitivity evaluation network and outputting the control response sensitivity coefficient.
[0010] In a possible implementation, adaptive optimization adjustment of the vehicle controller based on the controller performance evaluation results includes: obtaining the expected controller performance evaluation results; performing difference analysis on the controller performance evaluation results based on the expected controller performance evaluation results to obtain performance evaluation difference characteristics; and performing adaptive optimization adjustment of the vehicle controller based on the performance evaluation difference characteristics.
[0011] In one possible implementation, a controller performance warning signal is generated based on the performance evaluation difference characteristics.
[0012] This application also provides a controller performance optimization system for intelligent vehicles, the system comprising: a data acquisition module for acquiring vehicle operation data and vehicle control command data of a target vehicle; a state matrix construction module for constructing a controller state matrix based on the vehicle operation data; a performance evaluation module for performing multi-dimensional performance evaluation of the controller based on the vehicle control command data and the controller state matrix, and obtaining controller performance evaluation results; and a controller optimization module for adaptively optimizing and adjusting the vehicle controller based on the controller performance evaluation results.
[0013] This application proposes a method and system for optimizing the controller performance of intelligent vehicles. The method involves acquiring vehicle operation data and vehicle control command data of the target vehicle; constructing a controller state matrix based on the vehicle operation data; performing a multi-dimensional performance evaluation of the controller based on the vehicle control command data and the controller state matrix; and adaptively optimizing and adjusting the vehicle controller based on the performance evaluation results. This addresses the technical problems in existing technologies, such as the lack of a comprehensive, multi-dimensional evaluation system for controller performance and the difficulty in adaptively and dynamically optimizing the controller online based on the real-time operating status of the vehicle and the response to control commands. The goal is to improve the adaptive adjustment capability of the vehicle controller during actual operation, the comprehensiveness of the multi-dimensional performance evaluation, and the control quality and dynamic response accuracy throughout its entire lifecycle. Attached Figure Description
[0014] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments of this disclosure will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.
[0015] Figure 1 This is a schematic flowchart of a method for optimizing the controller performance of an intelligent vehicle, as provided in an embodiment of this application.
[0016] Figure 2 This is a schematic diagram of the controller performance optimization system structure for an intelligent vehicle provided in an embodiment of this application.
[0017] Figure labeling: Data acquisition module 10, state matrix construction module 20, performance evaluation module 30, controller optimization module 40. Detailed Implementation
[0018] To further illustrate the technical means and effects adopted by the present invention in order to achieve the intended purpose, the following detailed description is provided in conjunction with the accompanying drawings and preferred embodiments, based on the specific implementation methods, structures, features and effects of the present invention.
[0019] This application provides a method for optimizing the controller performance of intelligent vehicles, such as... Figure 1 As shown, the method includes: Step S100: Obtain vehicle operation data and vehicle control command data of the target vehicle.
[0020] Step S100 further includes acquiring the operating parameter acquisition results and control command acquisition results of the target vehicle; performing data validity verification and outlier removal on the operating parameter acquisition results to generate the vehicle operating data; and performing data validity verification and outlier removal on the control command acquisition results to generate the vehicle control command data.
[0021] Preferably, various sensors deployed on the target vehicle, such as vehicle speed sensors, inertial measurement units, wheel speed sensors, and acceleration sensors, are used to collect raw physical quantity data reflecting the vehicle's own motion state in real time. These data serve as the target vehicle's operating parameter collection results, including at least the vehicle's current longitudinal speed, lateral acceleration, yaw rate, engine speed, and wheel speed. Raw electrical signals or bus data from the vehicle controller, such as the vehicle control unit (VCU) or autonomous driving domain controller, used for actuators such as drive motors, braking systems, and steering systems, are also acquired as control command collection results. These data include at least the desired driving torque, desired braking pressure, and desired steering wheel angle.
[0022] Preferably, the collected operating parameters undergo data validity verification, which involves logical and physical rationality checks, including range verification, timestamp verification, and integrity verification. Range verification checks whether the data is within the reasonable range specified by the sensor; for example, the vehicle speed value cannot be negative or exceed the vehicle's physical limits. Timestamp verification checks whether the time stamps of the data packets are continuous and whether there are any dropped frames or time jumps. Integrity verification checks whether data frames are lost or damaged during transmission. Then, outlier removal is performed, which identifies and removes data points that deviate significantly from the true trend due to sensor failure, electromagnetic interference, or communication errors. For example, if a sudden speed jump occurs at a certain moment that is much larger than the rate of change before and after, this point is considered an outlier and removed. Finally, reliable vehicle operating data for analysis and calculation is generated.
[0023] Preferably, the data validity of the control command acquisition results is verified, i.e., a rationality check is performed, including command range verification, command consistency verification, and communication protocol verification. Among them, command range verification is used to check whether the control command exceeds the physical response limit of the actuator. For example, the expected braking pressure should not exceed the maximum pressure that the braking system can provide. Command consistency verification is used to check whether the command contradicts the control logic. For example, when the vehicle is in drive, there should be no expected reverse command. Communication protocol verification is used to check whether the data conforms to the communication protocol standard of the controller's local area network or Ethernet, such as whether the cyclic redundancy check is correct. Then, outlier removal is performed, i.e., removing erroneous command signals caused by communication interference or momentary controller failure. For example, if the expected torque command suddenly becomes a value far exceeding the normal range at a certain moment, the command value is removed. Finally, a set of commands that can truly reflect the controller's intention is generated.
[0024] Step S200: Construct a controller state matrix based on the vehicle operation data.
[0025] Step S200 further includes performing time alignment and data normalization processing on the vehicle operation data to generate a standardized vehicle operation data sequence; and performing matrix mapping on the standardized vehicle operation data sequence to generate the controller state matrix.
[0026] Preferably, due to the different sampling frequencies and data transmission delays of different sensors (such as vehicle speed sensors, inertial measurement units, and GPS), the collected various types of operational data are inconsistent on the time axis. The vehicle operational data after validity verification is time-aligned, and all heterogeneous data are rearranged or interpolated according to a unified reference time axis to ensure that all types of data correspond to the vehicle's motion state at the same time point in the same time section. Then, the time-aligned vehicle operational data is normalized, including mapping data of different dimensions and orders of magnitude to the same numerical range, such as [0, 1] or a distribution with a mean of 0 and a variance of 1, through min-max normalization or Z-score standardization, to eliminate the influence of dimensions. Finally, a standardized vehicle operational data sequence is generated. Then, the standardized vehicle operation data sequence is matrix-mapped, that is, the standardized vehicle operation data sequence is converted into a matrix structure according to preset rules. The rows of the matrix represent different time points or sampling times, and the columns of the matrix represent different types of vehicle operation parameters, such as vehicle speed, acceleration, yaw rate, etc., generating a controller state matrix. This matrix encapsulates the multi-dimensional operation state of the target vehicle within a time window in a structured way. Each row represents the vehicle state vector at a certain moment. The entire matrix constitutes a state space that represents the current working environment and response results of the controller.
[0027] Step S300: Perform a multi-dimensional performance evaluation of the controller based on the vehicle control command data and the controller state matrix, and obtain the controller performance evaluation results.
[0028] Step S300 further includes: constructing a vehicle control command matrix based on the vehicle control command data; evaluating the control response sensitivity based on the vehicle control command matrix and the controller state matrix to obtain a control response sensitivity coefficient; evaluating the control error based on the vehicle control command matrix and the controller state matrix to obtain a control error coefficient; evaluating the control stability based on the vehicle control command matrix and the controller state matrix to obtain a control stability coefficient; and outputting the control response sensitivity coefficient, the control error coefficient, and the control stability coefficient as the controller performance evaluation result.
[0029] Preferably, vehicle control command data containing desired torque, desired steering angle, and desired braking force is converted into a matrix structure. The rows of the matrix represent different time points or sampling times, and the columns represent different types of control commands, characterizing the control input sequence that the vehicle is expected to execute within a time window. Based on the vehicle control command data and the controller state matrix, a multi-dimensional performance evaluation of the controller is performed, including control response sensitivity evaluation, control error evaluation, and control stability evaluation. Specifically, the control response sensitivity evaluation assesses the controller's response speed and following ability to changes in commands, that is, how quickly the actual state quantities such as actual vehicle speed and actual steering angle in the controller state matrix follow the changes when the control command matrix undergoes a step or change. By comparing the change time of the command matrix with the response time of the state matrix, parameters such as response delay time and rise time are analyzed and comprehensively calculated to determine the control response sensitivity coefficient. The higher the control response sensitivity coefficient, the faster the controller responds to the command. The control error evaluation assesses the degree of deviation between the actual output of the controller and the desired command during steady-state or dynamic processes. The accuracy of the controller is determined by calculating the difference between the vehicle control command matrix (expected value) and the controller state matrix (actual value) element-by-element or time-by-time, analyzing statistical indicators such as absolute error, relative error, and integral error, and comprehensively determining the control error coefficient. The lower the control error coefficient, the higher the control accuracy of the controller, and the closer the actual value is to the expected value. The control stability evaluation assesses the fluctuation and convergence characteristics of the controller's output state during command execution, i.e., whether the control process is smooth and whether there are violent oscillations or divergences. Time series analysis is performed on the actual state quantities in the controller state matrix to calculate characteristic parameters such as variance, overshoot, and oscillation frequency, and then comprehensively determine the control stability coefficient. The higher the control stability coefficient, the smoother the control process and the smaller the fluctuation. Finally, the control response sensitivity coefficient, control error coefficient, and control stability coefficient are integrated to obtain and output a multi-dimensional controller performance evaluation result.
[0030] Furthermore, step S300 also includes: performing control response delay analysis based on the vehicle control command matrix and the controller state matrix to obtain control response delay parameters; training a control response sensitivity evaluation network based on historical control response sensitivity evaluation records; inputting the control response delay parameters into the control response sensitivity evaluation network to output the control response sensitivity coefficient.
[0031] Preferably, control response delay analysis refers to comparing the correspondence between the vehicle control command matrix and the controller state matrix on the time axis to analyze the specific value of the system response delay, that is, the time point when the control quantity in the detection command matrix changes and the time point when the corresponding physical quantity in the state matrix begins to change, and calculating the time difference between the two as the control response delay parameter, such as the delay time between the issuance of the desired torque command and the start of actual vehicle speed change; using a large amount of controller performance evaluation data accumulated in the past as training samples, a training evaluation model is constructed according to the neural network, where the training samples refer to historical control response sensitivity evaluation records, including input features in historical scenarios, such as various delay parameters, vehicle states, etc., and their corresponding evaluation labels, such as manually labeled or known sensitivity levels, so that the neural network can learn and establish a nonlinear mapping relationship between input features and sensitivity evaluation results, generating a control response sensitivity evaluation network; the control response delay parameter obtained by real-time analysis is input into the control response sensitivity evaluation network, and the input delay parameter is calculated and reasoned according to the mapping relationship learned in the training stage, outputting a comprehensive control response sensitivity coefficient.
[0032] Step S400: Adaptive optimization adjustment of the vehicle controller is performed based on the controller performance evaluation results.
[0033] Step S400 further includes: obtaining the expected result of controller performance evaluation; performing difference analysis on the controller performance evaluation result based on the expected result of controller performance evaluation to obtain performance evaluation difference characteristics; and performing adaptive optimization adjustment on the vehicle controller based on the performance evaluation difference characteristics.
[0034] Preferably, based on the standard performance parameters calibrated at the vehicle's factory, the target threshold dynamically adjusted under the current operating condition, or the optimal control index calculated through a theoretical model, the expected results of controller performance evaluation are obtained. This is a set of benchmark values representing the ideal or target performance of the controller, including the expected control response sensitivity coefficient, the expected control error coefficient, and the expected control stability coefficient. Then, the actual calculated controller performance evaluation results are compared and quantitatively analyzed dimensionally with the expected results. The difference, rate of change, or direction of deviation between the actual and expected values in each dimension is calculated. For example, if the actual response sensitivity coefficient is lower than the expected value, or the actual control error coefficient is higher than the expected value, the performance evaluation difference characteristics are obtained. The system quantifies the gap between the current controller performance and the target state, indicating the shortcomings and their degree. Finally, the performance evaluation difference characteristics are used as input for PID control optimization. The system calculates and determines the necessary adjustments to the control parameters inside the vehicle controller, such as PID coefficients, filter cutoff frequency, and feedforward gain, including the direction and degree of deviation indicated by the performance evaluation difference characteristics. The adjustment amount of the control parameters is calculated by means of table lookup, gradient descent, or model prediction. The calculated parameter adjustment amount is then written into the vehicle controller to update the controller's operating parameters online. This allows the actual performance of the target vehicle controller to approach the expected result in subsequent control cycles, thus forming a closed-loop adaptive optimization.
[0035] Furthermore, step S400 also includes generating a controller performance warning signal based on the performance evaluation difference characteristics.
[0036] Preferably, after acquiring the performance evaluation difference characteristics, threshold judgment or trend analysis is performed. When the performance difference is detected to exceed the preset safety range or normal fluctuation range, a signal is generated to indicate abnormal controller performance. Specifically, based on the historical data of the vehicle controller, the maximum allowable deviation of the controller response sensitivity, error coefficient, and stability coefficient from the expected result is preset to determine the warning threshold. Then, the currently calculated performance evaluation difference characteristics are compared with the warning threshold to determine whether the absolute value of the performance difference exceeds the upper limit threshold, whether the combination of multiple performance differences meets the warning conditions, and whether the rate of change of the performance difference is too fast. When the judgment conditions are met, a digital signal or logic flag is generated to represent the controller performance warning signal and output to the vehicle display unit to warn the driver. At the same time, it is uploaded to the remote monitoring platform to record fault information and notify maintenance personnel.
[0037] In the above text, refer to Figure 1 A method for optimizing the controller performance of an intelligent vehicle according to an embodiment of the present invention is described in detail. Next, reference will be made to... Figure 2 A controller performance optimization system for an intelligent vehicle according to an embodiment of the present invention is described.
[0038] The intelligent vehicle controller performance optimization system according to embodiments of the present invention addresses the technical problems in the prior art, namely, the lack of a comprehensive, multi-dimensional evaluation system for controller performance and the difficulty in adaptively and dynamically optimizing the controller online based on the real-time operating status of the vehicle and the response to control commands. It achieves the technical effects of improving the adaptive adjustment capability of the vehicle controller during actual operation, the comprehensiveness of multi-dimensional performance evaluation, and the control quality and dynamic response accuracy throughout its entire lifecycle. Figure 2 As shown, the intelligent vehicle controller performance optimization system includes: a data acquisition module 10, a state matrix construction module 20, a performance evaluation module 30, and a controller optimization module 40.
[0039] The data acquisition module 10 is used to acquire vehicle operation data and vehicle control command data of the target vehicle; the state matrix construction module 20 is used to construct a controller state matrix based on the vehicle operation data; the performance evaluation module 30 is used to perform multi-dimensional performance evaluation of the controller based on the vehicle control command data and the controller state matrix, and obtain the controller performance evaluation results; the controller optimization module 40 is used to adaptively optimize and adjust the vehicle controller based on the controller performance evaluation results.
[0040] The specific configuration of the data acquisition module 10 will be described in detail below. The data acquisition module 10 further includes: acquiring the operating parameter acquisition results and control command acquisition results of the target vehicle; performing data validity verification and outlier removal on the operating parameter acquisition results to generate the vehicle operating data; and performing data validity verification and outlier removal on the control command acquisition results to generate the vehicle control command data.
[0041] The specific configuration of the state matrix construction module 20 will be described in detail below. The state matrix construction module 20 further includes: performing time alignment and data normalization processing on the vehicle operation data to generate a standardized vehicle operation data sequence; and performing matrix mapping on the standardized vehicle operation data sequence to generate the controller state matrix.
[0042] The specific configuration of the performance evaluation module 30 will be described in detail below. The performance evaluation module 30 further includes: constructing a vehicle control command matrix based on the vehicle control command data; evaluating the control response sensitivity based on the vehicle control command matrix and the controller state matrix to obtain a control response sensitivity coefficient; evaluating the control error based on the vehicle control command matrix and the controller state matrix to obtain a control error coefficient; evaluating the control stability based on the vehicle control command matrix and the controller state matrix to obtain a control stability coefficient; and outputting the control response sensitivity coefficient, the control error coefficient, and the control stability coefficient as the controller performance evaluation result.
[0043] The specific configuration of the performance evaluation module 30 will be described in detail below. The performance evaluation module 30 further includes: performing control response delay analysis based on the vehicle control command matrix and the controller state matrix to obtain control response delay parameters; training a control response sensitivity evaluation network based on historical control response sensitivity evaluation records; inputting the control response delay parameters into the control response sensitivity evaluation network; and outputting the control response sensitivity coefficient.
[0044] The specific configuration of the controller optimization module 40 will be described in detail below. The controller optimization module 40 further includes: acquiring expected controller performance evaluation results; performing difference analysis on the controller performance evaluation results based on the expected controller performance evaluation results to acquire performance evaluation difference characteristics; and performing adaptive optimization adjustments on the vehicle controller based on the performance evaluation difference characteristics.
[0045] The specific configuration of the controller optimization module 40 will be described in detail below. The controller optimization module 40 further includes: generating a controller performance early warning signal based on the performance evaluation difference characteristics.
[0046] The intelligent vehicle controller performance optimization system provided in this embodiment of the invention can execute the intelligent vehicle controller performance optimization method provided in this embodiment of the invention, and has the corresponding functional modules and beneficial effects of executing the method.
[0047] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.
Claims
1. A method for optimizing the controller performance of an intelligent vehicle, characterized in that, The method includes: Acquire vehicle operation data and vehicle control command data of the target vehicle; Based on the vehicle operation data, construct the controller state matrix; Based on the vehicle control command data and the controller state matrix, a multi-dimensional performance evaluation of the controller is performed to obtain the controller performance evaluation results. The vehicle controller is adaptively optimized and adjusted based on the controller performance evaluation results.
2. The method for optimizing the controller performance of an intelligent vehicle as described in claim 1, characterized in that, Acquire vehicle operation data and vehicle control command data of the target vehicle, including: Acquire the operating parameters and control commands of the target vehicle. The collected operating parameters are validated for data validity and outlier removal is performed to generate the vehicle operating data. The data validity of the control command acquisition results is verified and outliers are removed to generate the vehicle control command data.
3. The method for optimizing the controller performance of an intelligent vehicle as described in claim 1, characterized in that, Based on the vehicle operation data, a controller state matrix is constructed, including: The vehicle operation data is time-aligned and data-normalized to generate a standardized vehicle operation data sequence. The standardized vehicle operation data sequence is matrix-mapped to generate the controller state matrix.
4. The method for optimizing the controller performance of an intelligent vehicle as described in claim 1, characterized in that, Based on the vehicle control command data and the controller state matrix, a multi-dimensional performance evaluation of the controller is performed to obtain the controller performance evaluation results, including: Based on the vehicle control command data, construct a vehicle control command matrix; The control response sensitivity is evaluated based on the vehicle control command matrix and the controller state matrix to obtain the control response sensitivity coefficient. The control error is evaluated based on the vehicle control command matrix and the controller state matrix to obtain the control error coefficient. The control stability is evaluated based on the vehicle control command matrix and the controller state matrix to obtain the control stability coefficient. The control response sensitivity coefficient, the control error coefficient, and the control stability coefficient are output as the controller performance evaluation results.
5. The method for optimizing the controller performance of an intelligent vehicle as described in claim 4, characterized in that, The control response sensitivity is evaluated based on the vehicle control command matrix and the controller state matrix to obtain the control response sensitivity coefficient, including: The control response delay is analyzed based on the vehicle control command matrix and the controller state matrix to obtain the control response delay parameters; Train the control response sensitivity evaluation network based on historical control response sensitivity evaluation records; The control response delay parameter is input into the control response sensitivity evaluation network, and the control response sensitivity coefficient is output.
6. The method for optimizing the controller performance of an intelligent vehicle as described in claim 1, characterized in that, Based on the controller performance evaluation results, the vehicle controller is adaptively optimized and adjusted, including: Obtain the expected results of controller performance evaluation; Based on the expected results of the controller performance evaluation, the differences in the controller performance evaluation results are analyzed to obtain the performance evaluation difference characteristics. The vehicle controller is adaptively optimized and adjusted based on the performance evaluation differences.
7. The method for optimizing the controller performance of an intelligent vehicle as described in claim 6, characterized in that, Based on the performance evaluation differences, a controller performance warning signal is generated.
8. A controller performance optimization system for intelligent vehicles, characterized in that, The system is used to implement the controller performance optimization method for intelligent vehicles according to any one of claims 1 to 7, the system comprising: The data acquisition module is used to acquire vehicle operation data and vehicle control command data of the target vehicle; The state matrix construction module is used to construct the controller state matrix based on the vehicle operation data; The performance evaluation module is used to perform multi-dimensional performance evaluation of the controller based on the vehicle control command data and the controller state matrix, and obtain the controller performance evaluation results. The controller optimization module is used to adaptively optimize and adjust the vehicle controller based on the controller performance evaluation results.