Optimization method based on intelligent driving assistance system, electronic device, and storage medium

Abstract

The application discloses an optimization method based on an intelligent driving assistance system, an electronic device and a storage medium. The application inputs the obtained actual road condition data into the intelligent driving assistance system for simulation to obtain simulation driving data; performs multi-condition driving scoring on actual driving data corresponding to the actual road condition data and the simulation driving data to obtain multi-dimensional driving performance scores in multiple conditions; and optimizes the intelligent driving assistance system based on the multi-dimensional driving performance scores. Thus, the multi-dimensional driving performance scores obtained based on the actual driving data and the simulation driving data can more accurately evaluate the performance of the intelligent driving assistance system, thereby improving the accuracy of optimization of the intelligent driving assistance system.

Description

Technical Field

[0001] This application relates to the field of intelligent technology, and in particular to an optimization method, electronic device and storage medium based on an intelligent driving assistance system. Background Technology

[0002] With the increasing prevalence of automotive driver assistance systems (ADAS), user demands for these systems are gradually rising, shifting from "functional control" to "good comfort." This shift in demand is driving ADAS to evolve from its initial focus on sensor input, perception processing, decision-making, planning, and optimization based on intelligent driving assistance systems to an end-to-end iterative upgrade. Therefore, the ability to directly, quickly, and effectively assess and optimize the performance of intelligent driving assistance systems has become a crucial aspect of current driver assistance testing.

[0003] Current optimization methods for intelligent driving assistance systems typically rely on the vehicle's subjective evaluation of the system. This approach is affected by objective conditions such as the driver's mental state, resulting in insufficient accuracy. Summary of the Invention

[0004] The main technical problem addressed by this application is to provide an optimization method, electronic device, and storage medium based on an intelligent driving assistance system, which can improve the accuracy of optimizing the intelligent driving assistance system.

[0005] To address the aforementioned technical problems, this application provides an optimization method based on an intelligent driving assistance system. The method is applied to a vehicle equipped with an intelligent driving assistance system. The method includes: inputting acquired actual road condition data into the intelligent driving assistance system for simulation to obtain simulated driving data; performing multi-condition driving scoring based on the actual driving data corresponding to the actual road condition data and the simulated driving data to obtain multi-dimensional driving performance scores under multiple conditions; and optimizing the intelligent driving assistance system based on the multi-dimensional driving performance scores.

[0006] In one embodiment, the step of performing multi-condition driving scoring based on actual driving data corresponding to the actual road condition data and simulated driving data to obtain multi-dimensional driving performance scores under multiple conditions includes: constructing a multi-dimensional evaluation model based on the actual driving data; and performing driving scoring on the simulated driving data based on the multi-dimensional evaluation model to obtain the multi-dimensional driving performance scores.

[0007] In one embodiment, the step of constructing a multi-dimensional evaluation model based on the actual driving data includes: classifying the actual driving data according to driving conditions to obtain driving sub-data under multiple driving conditions; extracting features from the driving sub-data under each driving condition to obtain actual feature values ​​under each driving condition; and constructing a multi-dimensional evaluation model for each driving condition based on the actual feature values ​​under each driving condition.

[0008] In one embodiment, the step of performing multi-condition driving scoring based on actual driving data corresponding to the actual road condition data and simulated driving data to obtain multi-dimensional driving performance scores under multiple conditions includes: extracting features from the actual driving data to obtain actual feature values ​​under multiple conditions; extracting features from the simulated driving data to obtain simulated feature values ​​under multiple conditions; obtaining multi-dimensional feature deviations under each condition based on each actual feature value and each simulated feature value; and determining the multi-dimensional driving performance scores under multiple conditions based on the multi-dimensional feature deviations under each condition.

[0009] In one embodiment, the multi-dimensional feature deviation includes feature deviations of multiple evaluation dimensions. The step of obtaining the multi-dimensional feature deviations under each working condition based on each actual feature value and each simulated feature value includes: setting multiple evaluation dimensions for each working condition; and obtaining the feature deviations of each evaluation dimension according to each simulated feature value and each actual feature value.

[0010] In one embodiment, the step of optimizing the intelligent driving assistance system based on the multi-dimensional driving performance score includes: determining a target optimization dimension based on the multi-dimensional driving performance score; determining a target optimization strategy for the intelligent driving assistance system based on the target optimization dimension; and optimizing the intelligent driving assistance system based on the target optimization strategy.

[0011] In one embodiment, the step of determining the target optimization dimension based on the multi-dimensional driving performance score includes: determining a target driving performance score that is less than a preset performance score threshold from the multi-dimensional driving performance scores; and determining the evaluation dimension corresponding to the target driving performance score that is less than the preset performance score threshold as the target optimization dimension.

[0012] In one embodiment, the step of determining the target optimization strategy of the intelligent driving assistance system based on the target optimization dimension includes: in response to the target optimization dimension being a timing dimension, obtaining the distance difference between the distance to the vehicle in front and a preset reference distance based on the simulated driving data; if the distance difference is greater than a preset distance threshold, then determining the optimization intervention timing as the target optimization strategy.

[0013] To address the aforementioned technical problems, this application provides an electronic device, including a memory and a processor. The memory stores program instructions, and the processor retrieves the program instructions from the memory to execute the aforementioned optimization method based on an intelligent driving assistance system.

[0014] To address the aforementioned technical problems, this application provides a computer-readable storage medium, comprising: storing program data, wherein the program data, when executed by a processor, is used to implement the aforementioned optimization method based on an intelligent driving assistance system.

[0015] The above scheme involves inputting acquired real-world road condition data into an intelligent driving assistance system for simulation, obtaining simulated driving data. Based on the actual driving data and simulated driving data corresponding to the real-world road condition data, multi-condition driving scores are calculated to obtain multi-dimensional driving performance scores under multiple conditions. The intelligent driving assistance system is then optimized based on these multi-dimensional driving performance scores. Therefore, obtaining multi-dimensional driving performance scores based on both actual and simulated driving data allows for a more accurate evaluation of the intelligent driving assistance system's performance, thereby improving the accuracy of intelligent driving assistance system optimization. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 This is a flowchart illustrating an exemplary embodiment of the optimization method based on an intelligent driving assistance system shown in this application; Figure 2 yes Figure 1 A flowchart illustrating an exemplary embodiment of S120 in the optimization method based on an intelligent driving assistance system is shown. Figure 3 yes Figure 1 A flowchart illustrating an exemplary embodiment of S130 in the optimization method based on an intelligent driving assistance system is shown. Figure 4 This is a block diagram illustrating an optimization device based on an intelligent driving assistance system, as shown in an exemplary embodiment of this application. Figure 5 This is a schematic diagram of the structure of an embodiment of the electronic device provided in this application; Figure 6 This is a schematic diagram of an embodiment of the computer-readable storage medium provided in this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It is understood that the specific embodiments described herein are only for explaining this application and not for limiting it. Furthermore, it should be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all structures. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] First, it's important to note that with the increasing prevalence of driver assistance systems (ADAS), user demands for these systems are rising, shifting from "functional control" to "good comfort." This shift in demand is driving ADAS to evolve from its initial focus on sensor input, perception processing, decision-making, planning, and optimization based on intelligent driving assistance systems to an end-to-end iterative upgrade. Therefore, directly, quickly, and effectively assessing and optimizing the performance of intelligent driving assistance systems has become a crucial task in current driver assistance testing. Currently, optimization methods for intelligent driving assistance systems typically rely on subjective evaluations of the system from actual vehicles. This approach is affected by objective conditions such as the driver's mental state, resulting in insufficient accuracy.

[0019] Based on this, this application provides an optimization method, electronic device, and computer-readable storage medium based on an intelligent driving assistance system. For details, please refer to [link / reference needed]. Figure 1 , Figure 1 This is a flowchart illustrating an exemplary embodiment of an optimization method based on an intelligent driving assistance system as shown in this application.

[0020] The execution entity of an optimization method based on an intelligent driving assistance system can be a terminal device, a server, or other processing device. The terminal device can be a computer, mobile device, terminal, computing device, in-vehicle device, etc. The execution entity of the optimization method based on an intelligent driving assistance system can also be an optimization device based on the intelligent driving assistance system. In some possible implementations, the optimization method based on the intelligent driving assistance system can be implemented by a processor calling computer-readable instructions stored in memory. The execution entity of the optimization method based on an intelligent driving assistance system can also be a big data cluster. A big data cluster is a computer system architecture formed by multiple computers connected through a network. The big data cluster can be deployed on a private cloud built with K8S (Kubernetes, a container orchestration engine).

[0021] Specifically, the optimization method based on an intelligent driving assistance system in this embodiment is applied to a vehicle equipped with an intelligent driving assistance system. The optimization method based on the intelligent driving assistance system includes the following steps: S110. Input the acquired actual road condition data into the intelligent driving assistance system for simulation to obtain simulated driving data.

[0022] Actual road condition data refers to data that reflects actual road conditions. Specifically, optimization devices based on intelligent driving assistance systems collect road condition data of the expected road using sensors to obtain actual road condition data. This actual road condition data includes road length, road width, road curvature, etc.

[0023] Intelligent driving assistance systems refer to systems used to assist in driving a vehicle.

[0024] Simulated driving data refers to the data output by intelligent driving assistance systems used to assist driving. Simulated driving data can include simulated vehicle speed, simulated acceleration, simulated steering wheel angle, simulated distance to the vehicle in front, simulated throttle torque, etc.

[0025] The optimization device based on the intelligent driving assistance system inputs the acquired actual road condition data into the intelligent driving assistance system for simulation to obtain simulated driving data.

[0026] In some embodiments, the optimization device based on the intelligent driving assistance system takes the collected actual road condition data as input and feeds it back into the intelligent driving assistance system through the SIL (Software-in-the-loop) simulation system to obtain simulated driving data.

[0027] It can be seen that the simulation-based recharge process significantly reduces the frequency and cost of real-vehicle testing.

[0028] S120. Based on actual driving data and simulated driving data corresponding to actual road conditions, multi-condition driving scores are obtained to obtain multi-dimensional driving performance scores under multiple conditions.

[0029] Real-world driving data refers to data collected by drivers during actual driving on real roads. Specifically, the optimization devices based on intelligent driving assistance systems collect data on drivers' driving in the most comfortable manner on real roads. This real-world driving data includes vehicle speed, acceleration, steering wheel angle, distance to the vehicle in front, and throttle torque, among other parameters.

[0030] Multi-dimensional driving performance rating refers to a driving score based on multiple evaluation dimensions. These dimensions may include: timing, amplitude, stability, accuracy, speed appropriateness, smoothness, agility, and trajectory regularity.

[0031] Multiple operating conditions can include following vehicle deceleration, cornering tracking, lane changing, overtaking, deceleration when entering or leaving ramps, lane changing and exiting ramps, merging into the main road, deceleration when entering a curve, and acceleration when exiting a curve.

[0032] As an example, the optimization device based on the intelligent driving assistance system performs multi-condition driving scoring based on actual driving data and simulated driving data corresponding to actual road conditions, and obtains multi-dimensional driving performance scores under multiple conditions. The steps include: constructing a multi-dimensional evaluation model based on actual driving data; and performing driving scoring on simulated driving data based on the multi-dimensional evaluation model to obtain multi-dimensional driving performance scores.

[0033] The steps of the optimization device based on the intelligent driving assistance system to build a multi-dimensional evaluation model based on actual driving data include: classifying the actual driving data into different working conditions to obtain driving sub-data under multiple working conditions; extracting features from the driving sub-data under each working condition to obtain actual feature values ​​under each working condition; and building a multi-dimensional evaluation model for each working condition based on the actual feature values ​​under each working condition.

[0034] The optimization device based on the intelligent driving assistance system divides the actual driving data within a preset time period into several windows according to the duration. Based on the actual driving data in each window, it obtains the working condition switching point, and divides the actual driving data into multiple segments according to each working condition switching point to obtain the driving sub-data under each working condition.

[0035] The optimization device based on the intelligent driving assistance system divides the actual driving data into several windows according to the duration, and calculates the vehicle speed change rate, longitudinal acceleration, lateral acceleration, steering wheel angle, and steering angle change rate for each window. When the vehicle speed change rate is greater than a preset change rate threshold and the duration is greater than a preset operating condition duration, the moment is determined as the deceleration operating condition switching point. When the lateral acceleration is greater than a preset lane change threshold and the steering wheel angle change rate is greater than a preset steering angle threshold, the moment is determined as the lane change operating condition switching point. According to each operating condition switching point, the continuous actual driving data is divided into multiple segments to obtain the driving sub-data of the operating condition corresponding to the operating condition switching point.

[0036] In one embodiment, the optimization device based on the intelligent driving assistance system traverses the actual driving data in a sliding window with a fixed duration of T=2s, and calculates the vehicle speed change rate Δv / Δt, longitudinal acceleration a_x, lateral acceleration a_y, steering wheel angle θ, and steering angle change rate dθ / dt within each window. The optimization device based on the intelligent driving assistance system marks the moment as a following deceleration switching point when Δv / Δt is greater than -0.5m / s² and lasts for more than 3s; it marks the moment as a lane change switching point when lateral acceleration a_y is greater than 0.3m / s² and steering wheel angle change rate is greater than a steering angle threshold; and it marks the moment as a curve tracking condition when lateral acceleration is greater than a preset acceleration threshold and steering wheel angle is greater than a preset angle threshold.

[0037] In another embodiment, the optimization device based on the intelligent driving assistance system traverses the actual driving data in a sliding window of fixed duration T. In each window, it calculates the vehicle speed change rate Δv / Δt, longitudinal acceleration a_x, lateral acceleration a_y, steering wheel angle θ, and the angle change rate dθ / dt. When the change of any of the above signals exceeds the corresponding preset threshold and the duration exceeds the minimum working condition duration, the moment is marked as the working condition switching point. The continuous data is divided into independent working condition segments with the switching point as the boundary to obtain each driving sub-data.

[0038] The optimization device based on the intelligent driving assistance system extracts features from the driving sub-data under various operating conditions to obtain the actual feature values ​​for each condition. For example, the actual feature values ​​for the following deceleration condition may include the vehicle speed at the start of deceleration, the distance to the vehicle in front at the start of deceleration, the distance-deceleration curve, the rate of change of deceleration, and the maximum deceleration. The optimization device based on the intelligent driving assistance system extracts these features from the driving sub-data under the following deceleration condition. Similarly, the actual feature values ​​for the cornering tracking condition may include the corner curvature, vehicle speed, steering wheel angle and its rate of change, the difference between the maximum and minimum steering wheel angles, lateral deviation, and the rate of change of lateral acceleration. The optimization device based on the intelligent driving assistance system extracts these features from the driving sub-data under the cornering tracking condition. For example, the actual characteristic values ​​under lane change conditions may include the rate of change of steering wheel angle, peak lateral acceleration, lane change completion time, lateral acceleration, and peak yaw rate. The optimization device based on the intelligent driving assistance system extracts the rate of change of steering wheel angle, peak lateral acceleration, lane change completion time, lateral acceleration, and peak yaw rate from the driving sub-data under lane change conditions.

[0039] The optimization device based on the intelligent driving assistance system constructs a multi-dimensional evaluation model for each operating condition based on the actual feature values ​​under each operating condition. Specifically, the optimization device marks the actual multi-dimensional driving score values ​​corresponding to the actual feature values ​​under each operating condition; it inputs the actual feature values ​​and their corresponding actual multi-dimensional driving score values ​​under each operating condition into the multi-dimensional evaluation model to be trained, thereby obtaining the predicted multi-dimensional driving score values ​​for each operating condition output by the multi-dimensional evaluation model; it calculates the loss value between the actual multi-dimensional driving score values ​​and the corresponding predicted multi-dimensional driving score values; it trains the multi-dimensional evaluation model to be trained with the goal of reducing the loss value, until the loss value meets the preset loss requirements, and the multi-dimensional evaluation model that meets the requirements is taken as the final multi-dimensional evaluation model.

[0040] The optimization device based on the intelligent driving assistance system scores simulated driving data using a multi-dimensional evaluation model, resulting in a multi-dimensional driving performance score. Specifically, the device categorizes the simulated driving data into different operating conditions, obtaining sub-data sets for each condition; it extracts features from each sub-data set to obtain simulation feature values ​​for each condition; and it inputs these feature values ​​into the multi-dimensional evaluation model to obtain the multi-dimensional driving performance score.

[0041] The simulation feature values ​​for following vehicle deceleration conditions can include the simulated vehicle speed at the start of deceleration, the distance to the preceding vehicle at the start of deceleration, the simulated distance-deceleration curve, the simulated maximum deceleration, and the simulated rate of change of deceleration. The optimization device based on the intelligent driving assistance system extracts these simulation feature values ​​from the simulated driving sub-data for following vehicle deceleration conditions. Similarly, the simulation feature values ​​for cornering tracking conditions can include the simulated corner curvature, simulated vehicle speed, simulated steering wheel angle and its rate of change, the difference between the maximum and minimum simulated steering wheel angles, simulated lateral deviation, and the simulated rate of change of lateral acceleration. The optimization device based on the intelligent driving assistance system extracts these simulation feature values ​​from the simulated driving sub-data for cornering tracking conditions. Furthermore, the simulation feature values ​​for lane changing conditions can include the simulated rate of change of steering wheel angle, the simulated peak lateral acceleration, the simulated lane change completion time, the simulated lateral acceleration, the simulated lane change trajectory, and the simulated peak yaw rate. The optimization device based on the intelligent driving assistance system extracts these simulation feature values ​​from the simulated driving sub-data for lane changing conditions.

[0042] The steps of constructing a multi-dimensional evaluation model for each working condition based on the optimization device of the intelligent driving assistance system may further include: the optimization device of the intelligent driving assistance system marking the actual values ​​of the multi-dimensional driving scores corresponding to the driving sub-data under each working condition; inputting the driving sub-data under each working condition and its corresponding actual values ​​of the multi-dimensional driving scores into the multi-dimensional evaluation model to be trained, respectively, to obtain the predicted values ​​of the multi-dimensional driving scores under each working condition output by the multi-dimensional evaluation model; calculating the loss value between the actual values ​​of the multi-dimensional driving scores and the corresponding predicted values ​​of the multi-dimensional driving scores; training the multi-dimensional evaluation model to be trained with the goal of reducing the loss value, until the loss value meets the preset loss requirements, and using the multi-dimensional evaluation model that meets the requirements as the final multi-dimensional evaluation model.

[0043] The optimization device based on the intelligent driving assistance system scores simulated driving data using a multi-dimensional evaluation model, resulting in a multi-dimensional driving performance score. Specifically, the optimization device classifies the simulated driving data into different operating conditions, obtaining simulated driving sub-data under multiple operating conditions; this sub-data is then input into the multi-dimensional evaluation model to obtain the multi-dimensional driving performance score.

[0044] As yet another example, based on the above embodiments, please refer to... Figure 2 , Figure 2 yes Figure 1 The diagram illustrates an exemplary embodiment of S120 in the optimization method based on an intelligent driving assistance system. Specifically, the process in S120 of performing multi-condition driving scoring based on actual driving data and simulated driving data corresponding to actual road conditions, and obtaining multi-dimensional driving performance scores under multiple conditions, includes the following steps: S210. Extract features from actual driving data to obtain actual feature values ​​under multiple operating conditions.

[0045] In one embodiment, the optimization device based on the intelligent driving assistance system classifies the actual driving data into driving sub-data under multiple driving conditions; and extracts features from the driving sub-data under each driving condition to obtain the actual feature values ​​under each driving condition.

[0046] S220. Extract features from the simulated driving data to obtain simulation feature values ​​under multiple operating conditions.

[0047] In one embodiment, the optimization device based on the intelligent driving assistance system classifies the simulated driving data into different operating conditions to obtain simulated driving sub-data under multiple operating conditions; and extracts features from the simulated driving sub-data under each operating condition to obtain the simulation feature values ​​under each operating condition.

[0048] S230. Obtain multi-dimensional feature deviations under various working conditions based on actual feature values ​​and simulation feature values.

[0049] Multidimensional feature deviation includes feature deviations of multiple evaluation dimensions. The steps of the optimization device based on the intelligent driving assistance system to obtain multidimensional feature deviations under each working condition based on each actual feature value and each simulated feature value include: setting multiple evaluation dimensions for each working condition; and obtaining the feature deviations of each evaluation dimension based on each simulated feature value and each actual feature value.

[0050] S240. Determine the multi-dimensional driving performance score under multiple working conditions based on the multi-dimensional characteristic deviations under each working condition.

[0051] In this system, the evaluation dimensions for each operating condition are the same. For example, the optimization device based on the intelligent driving assistance system calculates the feature deviation of each operating condition in the first evaluation dimension, and obtains the driving performance score for the first evaluation dimension based on the feature deviation of the first evaluation dimension; it calculates the feature deviation of each operating condition in the second evaluation dimension, and obtains the driving performance score for the second evaluation dimension based on the feature deviation of the second evaluation dimension; it calculates the feature deviation of each operating condition in the third evaluation dimension, and obtains the driving performance score for the third evaluation dimension based on the feature deviation of the third evaluation dimension. The first evaluation dimension is the timing dimension, the second evaluation dimension is the amplitude dimension, and the third evaluation dimension is the stability dimension.

[0052] When the evaluation dimensions for each driving condition are the same, the optimization device based on the intelligent driving assistance system performs weighted processing on the driving performance scores for each evaluation performance. Specifically, for the following deceleration condition: the weight of the timing dimension is k_1=0.3, the weight of the amplitude dimension is k_2=0.3, and the weight of the stability dimension is k_3=0.4; for the cornering tracking condition: the weight of the timing dimension is k_1=0.2, the weight of the amplitude dimension is k_2=0.3, and the weight of the stability dimension is k_3=0.5; for the lane changing condition: the weight of the timing dimension is k_1=0.2, the weight of the amplitude dimension is k_2=0.4, and the weight of the stability dimension is k_3=0.4.

[0053] The evaluation dimensions for each operating condition are not entirely the same. For example, the optimization device based on the intelligent driving assistance system sets timing, amplitude, and stability dimensions for the following deceleration condition.

[0054] The timing dimension is used to evaluate the deviation between the timing of simulated intervention and the actual intervention. The optimization device based on the intelligent driving assistance system calculates the difference between the simulated distance to the vehicle in front at the start of deceleration in the simulation feature values ​​and the actual distance to the vehicle in front at the start of deceleration in the actual feature values; this difference is determined as the feature deviation of the timing dimension.

[0055] For example, the feature bias in the timing dimension satisfies the following formula: Δd=|d_sim-d_ref| In the above formula, Δd represents the characteristic deviation of the timing dimension, d_sim represents the distance to the vehicle in front when the simulation starts to decelerate, and d_ref represents the distance to the vehicle in front when the simulation starts to decelerate.

[0056] The timing-based driving performance score follows the formula: Score_timing=max(0, 100-0.3×Δd / σ_d×100) In the above formula, Score_timing represents the driving performance score in the timing dimension, and σ_d represents the standard deviation of the deceleration start distance in multiple actual driving scenarios.

[0057] The amplitude dimension is used to evaluate the deviation between the amplitude of the simulated control quantity and the amplitude of the actual control quantity. The optimization device based on the intelligent driving assistance system calculates the difference between the simulated maximum deceleration in the simulation characteristics and the maximum deceleration in the actual characteristic values, and determines this difference as the characteristic deviation of the amplitude dimension.

[0058] For example, the feature bias in the magnitude dimension satisfies the following formula: Δa = |a_sim_max - a_ref_max| In the above formula, Δa represents the characteristic deviation of the amplitude dimension, a_sim_max represents the maximum deceleration in the simulation, and ref_max represents the maximum deceleration in the actual characteristic value.

[0059] The driving performance score in the amplitude dimension satisfies the following formula: Score_amplitude = max(0, 100 - 0.3×Δa / σ_a×100) In the above formula, Score_amplitude represents the driving performance score in the amplitude dimension, and σ_a represents the standard deviation of the actual maximum deceleration.

[0060] The stationarity dimension is used to evaluate the deviation between the simulated rate of change of control input and the actual rate of change of control input. An optimization device based on an intelligent driving assistance system calculates the difference between the root mean square of the simulated rate of change of deceleration and the root mean square of the actual rate of change of deceleration, and this difference is determined as the characteristic deviation of the stationarity dimension.

[0061] For example, the driving performance score for the stability dimension satisfies the following formula: Score_smoothness=max(0,100-0.4×|RMS_sim-RMS_ref| / σ_RMS×100) In the above formula, Score_smoothness represents the driving performance score of the smoothness dimension, RMS_sim represents the root mean square (RMS) of the simulated rate of change of deceleration, and RMS_ref represents the root mean square of the rate of change of deceleration in the actual eigenvalues.

[0062] The optimization device based on the intelligent driving assistance system also includes: weighting the driving performance scores of each evaluation dimension under the following deceleration condition to obtain the driving performance score under the following deceleration condition.

[0063] For example, the weight of the timing dimension is 0.3, the weight of the amplitude dimension is 0.3, and the weight of the stability dimension is 0.4. The driving performance score under the following deceleration condition satisfies the following formula: Score_total=0.3×Score_timing+0.3×Score_amplitude+0.4×Score_smoothness In the above formula, Score_total represents the driving performance score under the following vehicle deceleration condition.

[0064] Specifically, the optimization device based on the intelligent driving assistance system sets accuracy, stability, and speed rationality dimensions for the curve tracking condition.

[0065] The accuracy dimension is used to evaluate the accuracy with which a vehicle tracks a target path. An optimization device based on an intelligent driving assistance system calculates the difference between the root mean square of the simulated lateral deviation and the root mean square of the lateral deviation in the actual eigenvalues; this difference is then defined as the characteristic deviation of the accuracy dimension.

[0066] For example, the feature bias of the accuracy dimension satisfies the following formula: Δe_RMS = |RMS_sim-RMS_ref|.

[0067] In the above formula, Δe_RMS represents the characteristic deviation of the accuracy dimension, RMS_sim represents the root mean square of the simulation lateral deviation, and RMS_ref represents the root mean square of the lateral deviation in the actual characteristic value.

[0068] The driving performance score for the accuracy dimension satisfies the following formula: Score_accuracy=max(0,100-0.4×(Δe_RMS / σ_e_RMS)×100) In the above formula, Score_accuracy represents the driving performance score of the accuracy dimension, and σ_e_RMS represents the standard deviation of the lateral deviation RMS.

[0069] The stability dimension is used to evaluate the control smoothness of the steering wheel and vehicle body posture, avoiding oscillations. An optimization device based on the intelligent driving assistance system calculates the difference between the root mean square of the simulated lateral acceleration change rate and the preset benchmark root mean square of the lateral acceleration change rate in the actual eigenvalues; this difference is determined as the characteristic deviation of the stability dimension.

[0070] For example, the feature bias of the stability dimension satisfies the following formula: ΔJ_RMS=|RMS(da_y / dt)_sim-RMS(da_y / dt)_ref| In the above formula, ΔJ_RMS represents the characteristic deviation of the stability dimension, RMS(da_y / dt)_sim represents the root mean square of the simulated lateral acceleration rate of change da_y / dt, and RMS(da_y / dt)_ref represents the root mean square of the actual lateral acceleration rate of change in the eigenvalues.

[0071] The driving performance score for the stability dimension satisfies the following formula: Score_stability=max(0, 100-0.5×(ΔJ_RMS / σ_J_RMS)×100) In the above formula, Score_stability represents the driving performance score of the stability dimension, and σ_J_RMS represents the root mean square standard deviation of the rate of change of lateral acceleration in the actual eigenvalues.

[0072] The speed rationality dimension is used to evaluate whether the vehicle speed on curves matches the curvature, reflecting the trade-off between safety and efficiency. An optimization device based on an intelligent driving assistance system calculates the difference between the simulated vehicle speed and the actual speed in the feature values, and this difference is determined as the feature deviation of the speed rationality dimension.

[0073] For example, the feature deviation of the vehicle speed reasonableness dimension satisfies the following formula: Δv_rel=|v_sim-v_ref| / v_ref In the above formula, Δv_rel represents the characteristic deviation of the vehicle speed rationality dimension, v_sim represents the simulated vehicle speed, and v_ref represents the vehicle speed in the actual characteristic value.

[0074] The driving performance score for the speed rationality dimension satisfies the following formula: Score_speed = max(0, 100-25×Δv_rel) In the above formula, Score_speed represents the driving performance score of the vehicle speed reasonableness dimension.

[0075] The optimization device based on the intelligent driving assistance system also includes: weighting the driving performance scores of each evaluation dimension in the cornering tracking condition to obtain the driving performance score under the cornering tracking condition. For example, the weight of the accuracy dimension is 0.4, the weight of the stability dimension is 0.3, and the weight of the speed rationality dimension is 0.4. The driving performance score under the cornering tracking condition satisfies the following formula: Score_total_curve=0.4×Score_accuracy+0.3×Score_stability+ 0.3×Score_speed In the above formula, Score_total_curve represents the driving performance score under cornering tracking conditions.

[0076] Specifically, the optimization device based on the intelligent driving assistance system sets smoothness, agility, stability and trajectory regularity dimensions for lane change conditions.

[0077] The ride comfort dimension is used to evaluate the degree of impact and sway felt by occupants during lane changes. An optimization device based on an intelligent driving assistance system calculates the difference between the simulated peak lateral acceleration and the actual peak lateral acceleration values; this difference is determined as the characteristic deviation of the ride comfort dimension.

[0078] For example, the feature bias of the smoothness dimension satisfies the following formula: Δa_y_max = |a_y_max_sim - a_y_max_ref| In the above formula, Δa_y_max represents the characteristic deviation of the smoothness dimension, a_y_max_sim represents the peak value of the simulated lateral acceleration, and a_y_max_ref represents the peak value of the lateral acceleration in the actual characteristic value.

[0079] The driving performance score for the smoothness dimension satisfies the following formula: Score_comfort =max(0, 100 - 0.4×(Δa_y_max / σ_a_y_max)×100) In the above formula, Score_comfort represents the driving performance score of the comfort dimension, and σ_a_y_max represents the standard deviation of the peak lateral acceleration.

[0080] The agility dimension is used to evaluate the decisiveness and execution efficiency of lane-changing maneuvers. An optimization device based on an intelligent driving assistance system calculates the relative deviation between the simulated lane-changing completion time and the actual lane-changing completion time in the feature value, and this relative deviation is determined as the feature deviation of the agility dimension.

[0081] For example, the feature bias of the agility dimension satisfies the following formula: ΔT_rel =|T_LC_sim-T_LC_ref| / T_LC_ref In the above formula, ΔT_rel represents the characteristic deviation of the agility dimension, T_LC_sim represents the simulated lane change completion time, and T_LC_ref represents the lane change completion time in the actual characteristic value.

[0082] The agility dimension of driving performance is rated according to the following formula: Score_agility=max(0,100-30×ΔT_rel) In the above formula, Score_agility represents the driving performance score of the agility dimension. The agility dimension emphasizes time control and is more sensitive to deviations.

[0083] The stability dimension is used to evaluate the vehicle's ability to recover stability after a lane change, avoiding "dramatic" behavior. An optimization device based on an intelligent driving assistance system calculates the deviation between the decay rate or absolute integral of the simulated lateral acceleration oscillation and the corresponding actual lateral acceleration oscillation decay rate or absolute integral, and this deviation is determined as the characteristic deviation of the stability dimension.

[0084] For example, the optimization device based on the intelligent driving assistance system uses the deviation Δa_y_os between the first reverse peak a_y_overshoot of the lateral acceleration after returning to positive (after the first zero crossing) and the reverse peak in the actual eigenvalue as the feature deviation of the stability dimension.

[0085] The driving performance score for the stability dimension satisfies the following formula: Score_settling=max(0,100-0.5×(Δa_y_os / σ_a_y_os)×100) In the above formula, Score_settling represents the driving performance score of the stability dimension, and σ_a_y_os represents the standard deviation of the overshoot.

[0086] The trajectory normalization dimension is used to evaluate whether the lane change trajectory is smooth and meets expectations. An optimization device based on an intelligent driving assistance system calculates the maximum deviation between the simulated lane change trajectory and an ideal fifth-order polynomial trajectory, and uses this maximum deviation as the feature deviation of the trajectory normalization dimension.

[0087] The driving performance score for the trajectory normality dimension satisfies the following formula: Score_path=max(0,100-2.0×e_path_max×100) In the above formula, Score_path represents the driving performance score of the trajectory normalization dimension, and e_path_max represents the feature deviation of the trajectory normalization dimension. It is assumed that the lane width is 3.75m, and e_path_max is in meters. The trajectory normalization dimension is a deduction item with a low weight, but it is used to constrain abnormal trajectories.

[0088] The optimization device based on the intelligent driving assistance system also includes: weighting the driving performance scores of each evaluation dimension for lane change conditions to obtain the driving performance score for lane change conditions. For example, the weight of the smoothness dimension is 0.35, the weight of the agility dimension is 0.3, the weight of the stability dimension is 0.25, and the weight of the trajectory regularity dimension is 0.1. The driving performance score for lane change conditions satisfies the following formula: Score_total_LC=0.35xScore_comfort+0.30xScore_agility+0.25xScore_settling+0.10xScore_path In the above formula, Score_total_LC represents the driving performance score under lane change conditions.

[0089] As another example, the steps of performing multi-condition driving scores based on actual driving data and simulated driving data corresponding to actual road conditions to obtain multi-dimensional driving performance scores under multiple conditions include: scoring the actual driving data to obtain the actual driving performance score; scoring the simulated driving data to obtain the simulated driving performance score; and determining the multi-dimensional driving performance score by the difference between the actual driving performance score and the simulated driving performance score.

[0090] The multi-dimensional driving performance rating includes driving performance scores for each evaluation dimension. The steps for obtaining the actual driving performance score based on the optimization device of the intelligent driving assistance system to evaluate actual driving data include: classifying the actual driving data according to driving conditions to obtain driving sub-data for multiple driving conditions; extracting features from the driving sub-data for each driving condition to obtain actual feature values ​​for each driving condition; setting multiple evaluation dimensions for each driving condition and calculating the feature deviation of each evaluation dimension based on the actual feature values; and determining the actual driving performance score for each evaluation dimension based on the feature deviations.

[0091] For example, the actual characteristic value, characteristic deviation, and actual driving performance score satisfy the following formula: Score_i=max(0,100-k_i×|P_sim-P_ref| / σ_ref×100) Where Score_i is the actual driving performance score of the i-th evaluation dimension, k_i is the weight coefficient of the evaluation dimension, P_sim is the feature value of the actual driving data in the evaluation dimension, P_ref is the preset benchmark value, and σ_ref is the standard deviation of the benchmark data in the evaluation dimension.

[0092] The multi-dimensional driving performance rating includes driving performance scores for each evaluation dimension. The steps for obtaining the simulated driving performance score based on the optimization device of the intelligent driving assistance system to evaluate simulated driving data include: classifying the simulated driving data according to operating conditions to obtain simulated driving sub-data under multiple operating conditions; extracting features from the simulated driving sub-data under each operating condition to obtain simulation feature values ​​for each operating condition; setting multiple evaluation dimensions for each operating condition and calculating the feature deviation of each evaluation dimension based on the simulation feature values; and determining the simulated driving performance score for each evaluation dimension based on the feature deviations.

[0093] For example, simulation eigenvalues, eigenvalue deviations, and simulation driving performance scores satisfy the following formula: Score`_i= max(0,100- k_i×|P`_sim-P_ref| / σ_ref×100) Where Score`_i is the simulated driving performance score of the i-th evaluation dimension, k_i is the weight coefficient of the evaluation dimension, P`_sim is the simulation feature value of the simulated driving data in the evaluation dimension, P_ref is the preset benchmark value, and σ_ref is the standard deviation of the benchmark data in the evaluation dimension.

[0094] It can be seen that obtaining multi-dimensional driving performance scores based on actual driving data and simulated driving data enables more accurate evaluation of the performance of the intelligent driving assistance system, and establishes a working condition-level quantitative scoring standard based on the operation benchmark of excellent drivers, which can give specific quantitative scores for the performance of the intelligent driving assistance system under various working conditions, with objective and reproducible evaluation results.

[0095] S130. Optimize the intelligent driving assistance system based on the multi-dimensional driving performance score.

[0096] As an example, the optimization device based on the intelligent driving assistance system performs system optimization on the intelligent driving assistance system in response to the multi-dimensional driving performance score being greater than a preset allowable difference value; if the multi-dimensional driving performance score is less than or equal to the preset allowable difference value, the intelligent driving assistance system is not optimized. The system optimization may be initialization, restart and other operations.

[0097] As another example, the step of optimizing the intelligent driving assistance system by the optimization device based on the intelligent driving assistance system according to the multi-dimensional driving performance score includes: determining a target optimization dimension based on the multi-dimensional driving performance score; determining a target optimization strategy of the intelligent driving assistance system based on the target optimization dimension; and optimizing the intelligent driving assistance system based on the target optimization strategy.

[0098] The step of determining, by the optimization device based on the intelligent driving assistance system, the target optimization dimension based on the multi-dimensional driving performance score includes: determining a target driving performance score that is less than a preset performance score threshold from the multi-dimensional driving performance scores; and determining the evaluation dimension corresponding to the target driving performance score that is less than the preset performance score threshold as the target optimization dimension.

[0099] The preset performance score threshold may be 80. The optimization device based on the intelligent driving assistance system determines the dimension corresponding to the driving performance score less than 80 as the target optimization dimension, and outputs the difference root cause and feature deviation of the intelligent driving assistance system. The difference root cause refers to the fundamental reason for the difference between the simulated operation and the actual operation.

[0100] In one embodiment, the optimization device based on the intelligent driving assistance system determines whether the driving performance score of the timing dimension is less than a preset performance score threshold, and if so, determines that the target optimization dimension is the timing dimension; and calculates the feature deviation through the formula Δd=|d_sim-d_ref|. If d_sim>d_ref, the difference root cause of the timing dimension is calibrated as "early intervention", that is, the intervention timing is too early; if d_sim<d_ref, the difference root cause of the timing dimension is calibrated as "late intervention", that is, the intervention timing is too late.

[0101] In yet another embodiment, the optimization apparatus based on the intelligent driving assistance system determines whether the driving performance score of the amplitude dimension is less than a preset performance score threshold. If yes, the target optimization dimension is determined to be the amplitude dimension; and the characteristic deviation is calculated by the formula Δa=|a_sim_max-a_ref_max|. If a_sim_max>a_ref_max, the difference root cause of the amplitude dimension is calibrated as "over-control", and if a_sim_max<a_ref_max, the difference root cause of the amplitude dimension is calibrated as "under-control".

[0102] In still another embodiment, the optimization apparatus based on the intelligent driving assistance system determines whether the driving performance score of the stability dimension is less than a preset performance score threshold. If yes, the target optimization dimension is determined to be the stability dimension; and the difference root cause of the stability dimension is calibrated as "unsmooth control".

[0103] It should be noted that, in response to the presence of more than two target optimization dimensions, the optimization apparatus based on the intelligent driving assistance system outputs the target optimization dimensions in the order of driving performance scores from low to high, and outputs the difference root causes corresponding to at least two or more target optimization dimensions as composite root causes. The optimization apparatus based on the intelligent driving assistance system takes the difference root cause corresponding to the target optimization dimension with the lowest driving performance score as the main root cause, and takes the difference root causes corresponding to the remaining template optimization dimensions as secondary root causes.

[0104] The optimization apparatus based on the intelligent driving assistance system determines the target optimization strategy of the intelligent driving assistance system based on the target optimization dimension. As an example, the optimization apparatus based on the intelligent driving assistance system determines the target optimization strategy corresponding to the target optimization dimension from a preset optimization strategy mapping table. As another example, the optimization apparatus based on the intelligent driving assistance system determines the target optimization strategy corresponding to the target optimization dimension and the characteristic deviation from the preset optimization strategy mapping table. The preset optimization strategy mapping table stores the correspondence between preset optimization dimensions and preset optimization strategies, and also stores the correspondence between preset optimization dimensions, preset characteristic deviations and preset optimization strategies.

[0105] As Figure 3 shown, the optimization apparatus based on the intelligent driving assistance system obtains that the driving performance score of the timing dimension is 65 points, the difference root cause is "late intervention", and the characteristic deviation is 3m. The optimization apparatus based on the intelligent driving assistance system determines that the driving performance score of the timing dimension is less than 80 points, which indicates that the driving assistance system starts braking 3m later than the optimal control in actual operation, calibrates the difference root cause as "late intervention", records Δd, and determines the target optimization strategy from the preset optimization strategy mapping table. The target optimization strategy may be deceleration in advance.

[0106] The driving performance score in the amplitude dimension is 72 points, the root cause of difference is "over-control", and the feature deviation is 0.1g. Based on the optimization device of the intelligent driving assistance system, if the driving performance score in the amplitude dimension is <80 points, it indicates that the initial braking deceleration of the driving assistance system is 0.1g greater than the initial braking deceleration in actual operation. The root cause of difference is calibrated as "under-control", and Δa is recorded. The target optimization strategy is determined from the preset optimization strategy mapping table. The target optimization strategy can be to reduce the maximum deceleration limit.

[0107] The driving performance score for the stability dimension is 55 points, the root cause of the difference is "unsmooth control," and the characteristic deviation is 3 m / s³. Based on the optimization device of the intelligent driving assistance system, the driving performance score for the stability dimension is determined to be <80 points. This indicates that the deceleration change rate of the driving assistance system is 3 m / s³ greater than the actual deceleration change rate during operation. The root cause of the difference in the stability dimension is then identified as "unsmooth control." A target optimization strategy is determined from the preset optimization strategy mapping table. The target optimization strategy could be to optimize deceleration using a preset deceleration change rate smoothing strategy.

[0108] The optimization device based on the intelligent driving assistance system outputs each objective optimization dimension and a composite root cause consisting of the difference root causes corresponding to each objective optimization dimension, according to the driving performance score from low to high; and generates an evaluation report.

[0109] The optimization device based on the intelligent driving assistance system, after optimizing the intelligent driving assistance system, also includes: recording the current target optimization dimension, driving performance scores for each evaluation dimension, driving performance scores under each operating condition, root causes of differences, feature deviations, and target optimization strategies, and generating an evaluation report. The evaluation report also includes the version of the intelligent driving assistance system and its features, in order to track performance change trends of different versions of the intelligent driving assistance system.

[0110] The optimization device based on the intelligent driving assistance system also includes: weighting the driving performance scores of each evaluation dimension to obtain a comprehensive driving score.

[0111] For example, the overall driving score satisfies the following formula: Score_total=Σ(w_i×Score_i) In the above formula, Score_total represents the overall driving score, w_i represents the weight of the i-th evaluation dimension, and the weight allocation of each condition is determined according to the degree of influence of that dimension on driving comfort.

[0112] It can be seen that obtaining multi-dimensional driving performance scores based on actual driving data and simulated driving data can more accurately evaluate the performance of intelligent driving assistance systems, thereby improving the accuracy of intelligent driving assistance system optimization. Simultaneously, by defining the target optimization dimensions, the root causes of performance differences can be automatically located, rather than simply providing a vague "good / bad" judgment, thus providing precise direction for algorithm optimization. Furthermore, through weighted summation and multi-dimensional scoring, the evaluation results include both an overall score and detailed scores for each evaluation dimension, facilitating rapid problem identification and tracking of algorithm iteration effects.

[0113] Figure 4 This is a block diagram illustrating an optimization device based on an intelligent driving assistance system, as shown in an exemplary embodiment of this application. Figure 4 As shown, the exemplary optimization device 400 based on an intelligent driving assistance system includes: an acquisition module 410, a scoring module 420, and an optimization module 430. Specifically: The acquisition module 410 is used to input the acquired actual road condition data into the intelligent driving assistance system for simulation to obtain simulated driving data.

[0114] The scoring module 420 is used to perform multi-condition driving scoring based on actual driving data and simulated driving data corresponding to actual road conditions, and obtain multi-dimensional driving performance scores under multiple conditions.

[0115] Optimization module 430 is used to optimize the intelligent driving assistance system based on multi-dimensional driving performance scores.

[0116] In this exemplary optimization device based on an intelligent driving assistance system, acquired real-world road condition data is input into the intelligent driving assistance system for simulation to obtain simulated driving data. Multi-condition driving scores are then calculated based on the actual driving data and simulated driving data corresponding to the real-world road condition data, resulting in multi-dimensional driving performance scores under multiple conditions. The intelligent driving assistance system is then optimized based on these multi-dimensional driving performance scores. Therefore, obtaining multi-dimensional driving performance scores based on both actual and simulated driving data allows for a more accurate evaluation of the intelligent driving assistance system's performance, thereby improving the accuracy of intelligent driving assistance system optimization.

[0117] The functions of each module can be found in the implementation examples of the optimization method based on the intelligent driving assistance system, and will not be repeated here.

[0118] To implement the optimization method based on the intelligent driving assistance system described in the above embodiments, this application proposes another electronic device, please refer to [link / reference needed]. Figure 5 , Figure 5 This is a schematic diagram of the structure of an embodiment of the electronic device provided in this application.

[0119] Electronic device 500 includes memory 501 and processor 502, wherein memory 501 and processor 502 are coupled together.

[0120] The memory 501 is used to store program data, and the processor 502 is used to execute the program data to implement the optimization method based on the intelligent driving assistance system in the above embodiments.

[0121] In this embodiment, processor 502 can also be referred to as CPU (Central Processing Unit). Processor 502 may be an integrated circuit chip with signal processing capabilities. Processor 502 can also be a general-purpose processor, digital signal processor, application-specific integrated circuit, field-programmable gate array or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component. The general-purpose processor can be a microprocessor, or processor 502 can be any conventional processor.

[0122] This application also provides a computer-readable storage medium, such as Figure 6 As shown, the computer-readable storage medium 600 is used to store program data 601, which, when executed by the processor, is used to implement the optimization method based on the intelligent driving assistance system as described in the method embodiments of this application.

[0123] The methods involved in the optimization method embodiments of the intelligent driving assistance system in this application, when implemented as software functional units and sold or used as independent products, can be stored in a device, such as a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0124] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0125] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. The term "and / or" is merely a description of the association between related objects, indicating that three relationships can exist; for example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects are in an "or" relationship. Furthermore, "many" in this document means two or more. In addition, the term "at least one" in this document means any combination of at least two of any one or more of a plurality of elements, such as including at least one of A, B, and C, and may mean including any one or more elements selected from the set consisting of A, B, and C.

Claims

1. An optimization method based on an intelligent driving assistance system, characterized in that, The method is applied to a vehicle equipped with an intelligent driving assistance system, the method comprising: The acquired actual road condition data is input into the intelligent driving assistance system for simulation to obtain simulated driving data. Based on the actual driving data corresponding to the actual road conditions and the simulated driving data, a multi-condition driving score is obtained to obtain a multi-dimensional driving performance score under multiple conditions. The intelligent driving assistance system is optimized based on the multi-dimensional driving performance score.

2. The method according to claim 1, characterized in that, The step of performing multi-condition driving scoring based on the actual driving data corresponding to the actual road condition data and the simulated driving data to obtain multi-dimensional driving performance scores under multiple conditions includes: A multi-dimensional evaluation model is constructed based on the actual driving data. The driving performance score is obtained by scoring the simulated driving data based on the multi-dimensional evaluation model.

3. The method according to claim 2, characterized in that, The steps for constructing a multi-dimensional evaluation model based on the actual driving data include: The actual driving data is classified according to the driving conditions to obtain driving sub-data under multiple driving conditions; Feature extraction is performed on the driving sub-data under each working condition to obtain the actual feature values ​​under each working condition; A multi-dimensional evaluation model for each working condition is constructed based on the actual characteristic values ​​under each working condition.

4. The method according to claim 1, characterized in that, The step of performing multi-condition driving scoring based on the actual driving data corresponding to the actual road condition data and the simulated driving data to obtain multi-dimensional driving performance scores under multiple conditions includes: Feature extraction is performed on the actual driving data to obtain actual feature values ​​under multiple operating conditions; Feature extraction is performed on the simulated driving data to obtain simulation feature values ​​under multiple operating conditions; Multi-dimensional feature deviations under various working conditions are obtained based on each actual feature value and each simulation feature value. The multi-dimensional driving performance score under each working condition is determined based on the multi-dimensional feature deviation under each working condition.

5. The method according to claim 4, characterized in that, The multi-dimensional feature deviation includes feature deviations across multiple evaluation dimensions. The step of obtaining the multi-dimensional feature deviation under each working condition based on each actual feature value and each simulated feature value includes: Multiple evaluation dimensions are set for each working condition; The feature deviations of each evaluation dimension are obtained based on each simulated feature value and each actual feature value.

6. The method according to claim 1, characterized in that, The step of optimizing the intelligent driving assistance system based on the multi-dimensional driving performance score includes: The target optimization dimensions are determined based on the multi-dimensional driving performance score. The target optimization strategy of the intelligent driving assistance system is determined based on the target optimization dimension. The intelligent driving assistance system is optimized based on the aforementioned target optimization strategy.

7. The method according to claim 6, characterized in that, The step of determining the target optimization dimension based on the multi-dimensional driving performance score includes: Determine a target driving performance score that is less than a preset performance score threshold from the multi-dimensional driving performance scores; The evaluation dimension corresponding to the target driving performance score that is less than the preset performance score threshold is determined as the target optimization dimension.

8. The method according to claim 6, characterized in that, The step of determining the target optimization strategy of the intelligent driving assistance system based on the target optimization dimension includes: In response to the target optimization dimension being the timing dimension, the distance difference between the vehicle's distance to the vehicle in front and the preset benchmark distance is obtained based on the simulated driving data; If the distance difference is greater than a preset distance threshold, then the timing of the intervention will be determined as the target optimization strategy.

9. An electronic device, characterized in that, include: A memory and a processor, wherein the memory stores program instructions, and the processor retrieves the program instructions from the memory to perform the method as claimed in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, include: The system stores program data, which, when executed by a processor, is used to implement the method as described in any one of claims 1-8.

Images