Automatic driving performance optimization method, system and device, storage medium and product

Abstract

The invention discloses an automatic driving performance optimization method, system and device, a storage medium and a product, and relates to the technical field of automatic driving, and the method comprises the steps: obtaining a driving scene parameter and an automatic driving algorithm; constructing a simulation platform by using the driving scene parameters and an automatic driving algorithm, and performing a simulation test on the automatic driving system through the simulation platform to obtain a simulation test result; and adjusting the weight and the evaluation key point of the simulation test result through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of the automatic driving performance. According to the scheme, by introducing a multi-dimensional evaluation system, more accurate and more comprehensive performance evaluation of the automatic driving system can be realized, and one-sided evaluation caused by a single index in the past is avoided. Besides, by dynamically adjusting the weight of the simulation test result and the evaluation key point, more accurate and more comprehensive performance evaluation of the automatic driving system is realized, and the reliability and safety of the automatic driving system in a complex and changeable real driving environment are improved.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and particularly to methods, systems, devices, storage media, and products for optimizing the performance of autonomous driving. Background Art

[0002] When evaluating the performance of autonomous driving systems using current technologies, there are still many challenges. Traditional evaluation methods mostly rely on simple metrics such as the number of takeover times and cannot comprehensively reflect the performance of the system in complex road environments. In addition, a single test scenario often cannot cover the diversity and complexity of real driving. Therefore, there is an urgent need for a comprehensive evaluation method that can establish an effective linkage between virtual environments and real roads to achieve a more accurate and comprehensive evaluation of the performance of autonomous driving systems and improve the reliability and safety of autonomous driving systems in complex and changing real driving environments. Summary of the Invention

[0003] The main objective of this application is to provide a method, system, device, storage medium, and product for optimizing the performance of autonomous driving, aiming to solve the technical problem of how to improve the reliability and safety of autonomous driving systems in complex and changing real driving environments.

[0004] To achieve the above objective, this application proposes a method for optimizing the performance of autonomous driving, which includes:

[0005] Obtain driving scenario parameters and autonomous driving algorithms;

[0006] Construct a simulation platform using the driving scenario parameters and the autonomous driving algorithms, and conduct a simulation test on the autonomous driving system through the simulation platform to obtain a simulation test result;

[0007] Adjust the weights and evaluation focuses of the simulation test results through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of autonomous driving performance.

[0008] In one embodiment, the step of adjusting the weights and evaluation focuses of the simulation test results through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of autonomous driving performance includes:

[0009] Extract and analyze the features of the driving scenario parameters to obtain a scenario feature vector;

[0010] Calculate and adjust the weights of the multi-dimensional evaluation system according to the scenario feature vector;

[0011] Determine the evaluation focuses of the multi-dimensional evaluation system according to the adjusted weights of the multi-dimensional evaluation system and the evaluation results of each evaluation dimension, and generate a weighted comprehensive score;

[0012] Based on the weighted comprehensive score, the simulation test results are comprehensively evaluated with weights to obtain the evaluation results of the autonomous driving performance.

[0013] In one embodiment, before the step of determining the evaluation focus of the multi-dimensional evaluation system and generating the weighted comprehensive score according to the weights of the adjusted multi-dimensional evaluation system and the evaluation results of each evaluation dimension, it includes:

[0014] The simulation test results are scored through a pre-constructed multi-dimensional evaluation system to obtain the scoring results of each dimension;

[0015] A quantitative level evaluation is performed on the scoring results of each dimension to obtain the evaluation results of each evaluation dimension.

[0016] In one embodiment, the pre-constructed multi-dimensional evaluation system at least includes planning ability evaluation, game ability evaluation, and vehicle control proficiency evaluation;

[0017] The step of obtaining the initial evaluation results by scoring the simulation test results through a pre-constructed five-dimensional evaluation system includes:

[0018] For the planning ability evaluation, calculate the deviation between the planned path and the optimal path in the simulation test results, and the ratio of the driving time to the shortest driving time, and calculate the planning ability score based on the deviation and the ratio; and / or

[0019] For the game ability evaluation, identify the set of interaction events in the simulation test results, judge whether the decision of each event in the set of interaction events conforms to the preset rules to obtain the decision score, and calculate the game ability score based on the decision score; and / or

[0020] For the vehicle control proficiency evaluation, calculate the variance of the vehicle acceleration change rate and the mean square error of the vehicle deviating from the expected trajectory in the simulation test results, and calculate the vehicle control proficiency score based on the variance and the mean square error.

[0021] In one embodiment, the step of performing a quantitative level evaluation on the initial evaluation results to obtain the final evaluation results includes:

[0022] Perform quantitative calculation on the initial evaluation results to obtain the evaluation score;

[0023] Map the evaluation score into a preset quantitative level to obtain the evaluation results of each evaluation dimension.

[0024] In one embodiment, after the step of adjusting the weights and evaluation focus of the simulation test results through a pre-constructed multi-dimensional evaluation system to obtain the evaluation results of the autonomous driving performance, it further includes:

[0025] Optimize the autonomous driving algorithm based on the evaluation result of the autonomous driving performance;

[0026] Use the optimized autonomous driving algorithm to update the simulation platform.

[0027] In one embodiment, the step of using the optimized autonomous driving algorithm to update the simulation platform includes:

[0028] Utilize the optimized autonomous driving algorithm to adjust the parameters of the simulation platform according to the error between the simulation data and the real road test result, and update the simulation platform based on the adjusted parameters.

[0029] In one embodiment, the step of utilizing the optimized autonomous driving algorithm to adjust the parameters of the simulation platform according to the error between the simulation data and the real road test result includes:

[0030] Obtain the real road test result according to the optimized autonomous driving algorithm;

[0031] Compare whether the error between the simulation test result and the real road test result is greater than a preset threshold;

[0032] If so, adjust the parameters of the simulation platform according to the error between the simulation test result and the real road test result until the difference between the simulation test result and the real road test result is less than the preset threshold.

[0033] In addition, to achieve the above object, the present application also proposes an autonomous driving performance optimization system, and the autonomous driving performance optimization system includes:

[0034] An acquisition module, configured to acquire driving scenario parameters and an autonomous driving algorithm;

[0035] A simulation module, configured to construct a simulation platform by using the driving scenario parameters and the autonomous driving algorithm, and perform a simulation test on the autonomous driving system through the simulation platform to obtain a simulation test result;

[0036] An evaluation module, configured to adjust the weight and evaluation focus of the simulation test result through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of the autonomous driving performance.

[0037] In addition, to achieve the above object, the present application also proposes an autonomous driving performance optimization device, and the device includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the autonomous driving performance optimization method as described above.

[0038] In addition, to achieve the above object, the present application further provides a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the above-mentioned automatic driving performance optimization method are implemented.

[0039] In addition, to achieve the above object, the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps of the above-mentioned automatic driving performance optimization method are implemented.

[0040] The embodiments of the present application provide an automatic driving performance optimization method, system, device, storage medium and product. The method includes: obtaining driving scene parameters and an automatic driving algorithm; using the driving scene parameters and the automatic driving algorithm to construct a simulation platform, and performing a simulation test on the automatic driving system through the simulation platform to obtain a simulation test result; adjusting the weights and evaluation focuses of the simulation test result through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of the automatic driving performance. By introducing a multi-dimensional evaluation system, more accurate and comprehensive performance evaluation of the automatic driving system can be achieved, avoiding one-sided evaluation caused by a single index in the past. In addition, by dynamically adjusting the weights and evaluation focuses of the simulation test result, more accurate and comprehensive performance evaluation of the automatic driving system is realized, improving the reliability and safety of the automatic driving system in a complex and changeable real driving environment. Description of the Drawings

[0041] The drawings here are incorporated into the specification and form a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0042] To more clearly illustrate the technical solutions in the embodiments of the present application or in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0043] Figure 1 It is a schematic flowchart provided for the first embodiment of the automatic driving performance optimization method of the present application;

[0044] Figure 2 It is a schematic flowchart provided for the second embodiment of the automatic driving performance optimization method of the present application;

[0045] Figure 3 It is a schematic flowchart provided for the third embodiment of the automatic driving performance optimization method of the present application;

[0046] Figure 4It is a schematic flowchart of the autonomous driving performance optimization method provided by Embodiment 1 to Embodiment 3 of the present application;

[0047] Figure 5 It is a schematic diagram of the module structure of the autonomous driving performance optimization system according to the embodiment of the present application;

[0048] Figure 6 It is a schematic diagram of the device structure of the hardware operating environment involved in the autonomous driving performance optimization method according to the embodiment of the present application.

[0049] The implementation, functional features, and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. Detailed implementation manners

[0050] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0051] To better understand the technical solutions of the present application, the following will be described in detail with reference to the accompanying drawings of the specification and specific implementation manners.

[0052] The main solution of the embodiment of the present application is: obtaining driving scenario parameters and an autonomous driving algorithm; constructing a simulation platform by using the driving scenario parameters and the autonomous driving algorithm, and performing a simulation test on the autonomous driving system through the simulation platform to obtain a simulation test result; adjusting the weights and evaluation focuses of the simulation test result through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of the autonomous driving performance.

[0053] With the rapid development of autonomous driving technology, how to effectively evaluate the performance of autonomous driving systems has become the focus of the industry. Traditional evaluation methods mainly use the takeover times as the core indicator, that is, counting the number of times the driver needs to intervene or take over the vehicle during the test. However, this evaluation method has obvious limitations. First, the autonomous driving system may drive in an overly conservative manner. For example, it may choose to follow the vehicle when overtaking is possible, thereby reducing the takeover times but lowering the driving efficiency and unable to reflect the driving level of a skilled driver. Second, the takeover times cannot reflect the detailed performance of the autonomous driving system in aspects such as planning and decision-making, traffic game, and vehicle control. In addition, actual road tests are affected by many uncontrollable factors, such as traffic flow, weather conditions, and light changes, making the single test result have a large uncertainty. To solve the above problems, it is necessary to construct a more fine-grained evaluation system that comprehensively considers various capabilities of the autonomous driving system. Conduct tests in multiple different scenarios, increase the sample size, and reduce the influence of external variables to obtain a more objective and fair evaluation result. This need has promoted the research and application of fine-grained and multi-dimensional autonomous driving evaluation methods.

[0054] Currently, autonomous driving technology is developing rapidly. However, there are still many challenges in evaluating the performance of autonomous driving systems. Traditional evaluation methods mostly rely on simple metrics such as the number of takeovers and cannot comprehensively reflect the system's performance in complex road environments. In addition, a single test scenario often cannot cover the diversity and complexity of real driving.

[0055] This application provides a solution. By introducing a scenario adaptive adjustment mechanism into a multi-dimensional evaluation system, the weights of evaluation dimensions are dynamically adjusted for different test scenarios to achieve focused evaluation. This solution includes modules such as scenario feature analysis, adaptive adjustment of evaluation dimension weights, and adaptive evaluation. Through the linkage of simulation tests and real-road tests, a continuous feedback and optimization cycle is formed to improve the performance and reliability of the autonomous driving system in different scenarios.

[0056] It should be noted that the execution entity of this embodiment can be a computing service device with data processing, network communication, and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of implementing the above functions. Hereinafter, a personal computer is taken as an example to illustrate this embodiment and the following embodiments.

[0057] Based on this, the embodiments of this application provide an autonomous driving performance optimization method. Refer to Figure 1 , Figure 1 which is a schematic flowchart of the first embodiment of the autonomous driving performance optimization method of this application.

[0058] In this embodiment, the autonomous driving performance optimization method includes steps S10 to S30:

[0059] Step S10, obtain driving scenario parameters and an autonomous driving algorithm;

[0060] It should be noted that the driving scenario parameters include but are not limited to road type, traffic flow, weather conditions, lighting conditions, etc. The autonomous driving algorithm refers to the core algorithm in the driving system, which is responsible for vehicle decision-making, control, and execution.

[0061] Step S20, construct a simulation platform using the driving scenario parameters and the autonomous driving algorithm, and perform a simulation test on the autonomous driving system through the simulation platform to obtain a simulation test result;

[0062] It should be noted that the simulation platform is an integrated environment designed to simulate and test the performance of the autonomous driving system using the input driving scenario parameters and a specific autonomous driving algorithm.

[0063] Specifically, after selecting appropriate driving scenario parameters and autonomous driving algorithms, a corresponding virtual environment is constructed on the simulation platform, and this environment is used to test the selected driving scenario parameters and autonomous driving algorithms. During the test, the autonomous driving system is made to run in the virtual environment, and the system's reactions, decisions, and behaviors are recorded, including speed changes, path planning, obstacle avoidance measures, etc. The simulation platform can accurately simulate the physical rules of the real world, sensor characteristics, and vehicle dynamics, etc., to ensure the validity of the test results.

[0064] During the simulation process, the performance of the autonomous driving system is observed by changing different parameters. For example, the weather conditions are adjusted to check the system's reactions under different visibility conditions, or the traffic flow is increased to examine the system's processing ability.

[0065] Finally, the simulation test results output from the simulation platform are collected for subsequent evaluation of autonomous driving performance.

[0066] Through the above steps, the simulation test can not only provide a large amount of test data for the autonomous driving algorithm, but also provide evaluation parameters for the performance evaluation of the subsequent intelligent driving system.

[0067] Step S30, adjust the weights and evaluation focuses of the simulation test results through a pre-constructed multi-dimensional evaluation system to obtain the evaluation results of autonomous driving performance.

[0068] It should be noted that in this embodiment, the multi-dimensional evaluation system includes at least five dimensions, namely, planning ability, gaming ability, vehicle control proficiency, environmental perception ability, and human-machine interaction experience. The five dimensions are integrated into a comprehensive evaluation framework to form a five-dimensional evaluation system. The following embodiments are illustrated by taking the five-dimensional evaluation system as an example.

[0069] It is worth noting that due to the problems of one-sidedness and subjectivity in the existing autonomous driving evaluation methods for system performance evaluation, therefore, by performing step S30, the multi-faceted performance of the autonomous driving system can be comprehensively evaluated, avoiding the one-sided evaluation caused by the previous single index.

[0070] In a feasible embodiment, step S30 may further include steps S31 to S34:

[0071] Step S31, perform feature extraction and analysis on the driving scenario parameters to obtain a scenario feature vector;

[0072] It should be noted that in the optimization of autonomous driving performance, different driving scenarios have different requirements for the autonomous driving system. For example, in highway scenarios, planning ability and vehicle control proficiency may be crucial; while in complex urban road conditions, environmental perception ability and game-playing ability are particularly important. Therefore, this embodiment proposes a method for adaptively adjusting the key evaluation of the five-dimensional evaluation according to different scenarios.

[0073] Specifically, first, extract and analyze the features of the current driving scenario parameters. The current driving scenario parameters include, but are not limited to, road type (such as urban roads, highways), traffic flow (light, medium, heavy), weather conditions (sunny, rainy, snowy), lighting conditions (daytime, night), etc., to obtain the scenario feature vector S = {s1, s2,..., sn}, where the scenario feature vector is a vector composed of multiple elements, and each element represents a specific scenario feature. For example, s1 may be the encoded value of the road type, s2 is the quantified value of the traffic flow level, and so on. This vector comprehensively reflects the main features of the current driving scenario.

[0074] Step S32, calculate and adjust the weights of the multi-dimensional evaluation system according to the scenario feature vector;

[0075] In this embodiment, according to the scenario feature vector obtained in step B21, dynamically adjust the weights W = {w p , w b , w c , w e , w h} of the five evaluation dimensions.

[0076] Specifically, first, establish a mapping relationship between the scenario and the dimension. Through expert knowledge and data analysis, establish the association matrix M between different scenarios and evaluation dimensions. The element m ij of the matrix M represents the influence degree of the scenario feature vector si on the evaluation dimension. Among them, the association matrix M refers to a mathematical model used to quantify the influence degree of different driving scenario feature vectors si on each evaluation dimension.

[0077] Calculate the preliminary weights of the evaluation dimensions using the scenario feature vector and the association matrix:

[0078]

[0079] In the formula, S i represents the i-th scenario feature vector, and m ij represents the influence degree of the scenario feature vector si on the evaluation dimension.

[0080] Then, perform normalization processing on the preliminary weights:

[0081]

[0082] In the formula, w j is the preliminary weight.

[0083] Step S33: According to the weights of the adjusted multi-dimensional evaluation system and the evaluation results of each evaluation dimension, confirm the evaluation focus of the multi-dimensional evaluation system and generate a weighted comprehensive score;

[0084] It should be noted that the weighted comprehensive score refers to a single value that can comprehensively reflect the overall performance of the system when evaluating the autonomous driving system, based on the weights of the five-dimensional evaluation system adjusted according to different scenario feature vectors.

[0085] Specifically, according to the adjusted weight W, clarify which evaluation dimensions are more critical in the current scenario. For example, in an urban environment with complex and heavy traffic, the environmental perception ability and game-playing ability may be given higher weights; while on the highway, the planning ability and vehicle control proficiency may be more important.

[0086] Subsequently, using the adjusted weight W and the scores S of each dimension, calculate the weighted comprehensive score through the following formula:

[0087] S total = w p × S p + w b × S b + w c × S c + w e × S e + w h × S h

[0088] In the formula, w p represents the weight of the planning ability evaluation, w b represents the weight of the game-playing ability evaluation, w c represents the weight of the vehicle control proficiency evaluation, w e represents the weight of the environmental perception ability evaluation, w h represents the weight of the human-machine interaction experience evaluation, S p represents the score of the planning ability, S b represents the score of the game-playing ability, S c represents the score of the vehicle control proficiency evaluation, S e represents the score of the environmental perception ability, S h represents the score of the human-machine interaction experience.

[0089] When the scores of each evaluation dimension are multiplied by their corresponding weights and then summed up, a weighted comprehensive score that can reflect the overall performance of the autonomous driving system in a specific scenario is obtained. This score not only considers the absolute performance of the system in each dimension but also combines the relative importance of these dimensions in the current scenario, thus providing a more accurate and targeted evaluation result.

[0090] Step S34: Based on the weighted comprehensive score, conduct a weighted comprehensive evaluation of the simulation test results to obtain the evaluation result of the autonomous driving performance.

[0091] Specifically, when conducting a simulation test on the driving scenario parameters, in combination with the weighted comprehensive score, through the five-dimensional evaluation system with adjusted weights, conduct a weighted comprehensive evaluation of the simulation test results to obtain a more accurate and comprehensive evaluation result of the autonomous driving performance.

[0092] To better understand Steps S31 to S34, the following will illustrate with a specific scenario as an example.

[0093] For example, when the vehicle is automatically driving in a highway scenario, first, the system automatically or manually inputs the current driving scenario parameters, extracts and analyzes the scene feature vectors, and the scene feature vectors include but are not limited to road type (highway), traffic flow (medium), weather conditions (sunny), and lighting conditions (good). Then, according to the scene feature vectors, calculate the weights of the five-dimensional evaluation, including the weight (w p ) of the planning ability is a high weight, and actions such as overtaking and lane changing need to be planned; the weight (w c ) of the vehicle control proficiency is a high weight, and precise control of vehicle speed and distance is required; the weight (w e ) of the environmental perception ability is a medium weight, perceiving the vehicles in front and behind and road signs; the weight (w b ) of the game ability is a low weight, with less interaction on the highway; the weight (w h ) of the human-machine interaction experience is a low weight, with less driver intervention. Subsequently, calculate the weighted comprehensive score in combination with the weights of each dimension, that is, W = {w p : 0.4, w b : 0.1, w c : 0.3, w e : 0.15, w h : 0.05}. After the weights of the five-dimensional evaluation system are adjusted, when evaluating the driving scenario parameters and the autonomous driving algorithm, a more comprehensive and accurate evaluation result can be output.

[0094] For another example, when a vehicle is automatically driving in complex urban road conditions, first, the system automatically or manually inputs the current driving scenario parameters, analyzes and extracts the scenario feature vectors, which include but are not limited to road type (urban road), traffic flow (high), weather conditions (cloudy), and lighting conditions (average). Then, based on the scenario feature vectors, the weights of the five-dimensional evaluation are calculated, including the weight (w p ) of the planning ability is medium, and the path planning is complex; the weight (w c ) of the vehicle control proficiency is medium, and frequent starts and stops are required; the weight of the environmental perception ability (w e ) is high, and pedestrians need to be recognized; the weight of the game ability (w b ) is high, and frequent interactions with other vehicles are required; the weight of the human-machine interaction experience (w h ) is high, and the driver may need more information. Subsequently, the weighted comprehensive score is calculated by combining the weights of each dimension, that is, W = {w p : 0.2, w b : 0.3, w c : 0.2, w e : 0.25, w h : 0.05}. After the weight adjustment of the five-dimensional evaluation system, when evaluating the driving scenario parameters and the autonomous driving algorithm, a more comprehensive and accurate evaluation result can be output.

[0095] Through the above steps, on the basis of the original five-dimensional evaluation system, a scenario adaptive evaluation mechanism is introduced. This mechanism can automatically adjust the weights and evaluation focuses of each evaluation dimension according to the characteristics of the current test scenario, and realize targeted performance evaluation.

[0096] Through the method of the above embodiment, that is, obtaining the driving scenario parameters and the autonomous driving algorithm; constructing a simulation platform by using the driving scenario parameters and the autonomous driving algorithm, and performing simulation tests on the autonomous driving system through the simulation platform to obtain simulation test results; adjusting the weights and evaluation focuses of the simulation test results through a pre-constructed multi-dimensional evaluation system to obtain the evaluation results of the autonomous driving performance. This solution can achieve more accurate and comprehensive performance evaluation of the autonomous driving system by introducing a multi-dimensional evaluation system, and avoid the one-sided evaluation caused by a single index in the past. In addition, by dynamically adjusting the weights and evaluation focuses of the simulation test results, the accuracy and reliability of the performance evaluation of the autonomous driving system are improved.

[0097] Based on the first embodiment of the present application, in the second embodiment of the present application, the content that is the same as or similar to the above embodiment one can be referred to the above introduction, and will not be repeated hereinafter. On this basis, please refer to Figure 2, before step S33, the autonomous driving performance optimization method further includes steps S331 to S332:

[0098] Step S331, calculate the scores of the simulation test results through a pre - constructed multi - dimensional evaluation system to obtain the score results of each dimension;

[0099] It should be noted that the pre - constructed multi - dimensional evaluation system includes planning ability evaluation, game ability evaluation, vehicle control proficiency evaluation, environment perception ability, and human - machine interaction experience. The environment perception ability is usually quantitative, and the human - machine interaction experience is subjectively scored.

[0100] Specifically, for the planning ability evaluation, it includes path rationality evaluation and efficiency evaluation. First, calculate the deviation between the planned path data and the optimal path data in the simulation test results. Let Lp be the length of the planned path, Lopt be the length of the optimal path, and the deviation be ΔL = LP - Lopt. Then, calculate the ratio of the driving time Tp to the shortest driving time Topt Finally, comprehensively consider the deviation between the planned path and the optimal path and the ratio of the driving time to the shortest driving time to obtain the planning ability score S p = f(ΔL,E).

[0101] For the game ability evaluation, it includes interaction event detection, decision rationality evaluation, and score calculation. First, identify the set of interaction events {e1, e2,..., en} in the simulation test results, where ei is the i - th event in the set of interaction events. For each event ei, judge whether the decision of each event conforms to the preset rules (including traffic rules and safety criteria), and obtain the decision score si. Finally, calculate the game ability score where n is the total number of interaction events, and si is the decision score of the i - th interaction event.

[0102] For the vehicle control proficiency evaluation, it includes control smoothness evaluation, control accuracy evaluation, and score calculation. First, evaluate the vehicle control smoothness, including calculating the variance of the vehicle acceleration change rate a(t) Then, evaluate the vehicle control accuracy, including calculating the mean square error of the vehicle's deviation from the expected trajectory where y(t) is the actual trajectory within the entire test time T, and y d (t) is the expected trajectory within the entire test time T. Finally, comprehensively consider the smoothness and accuracy to obtain the vehicle control proficiency score

[0103] For the environment perception ability, it determines how the vehicle understands the surrounding world and makes decisions accordingly. The evaluation of this ability is usually quantitative, meaning it is based on measurable data and performance indicators.

[0104] For the human - machine interaction experience, it determines how the vehicle understands the surrounding world and makes decisions accordingly. The evaluation of this ability is usually quantitative, and users can rate the experience of autonomous driving through the interface.

[0105] By introducing a five - dimensional evaluation system, a more accurate and comprehensive evaluation of the performance of the autonomous driving system can be achieved. This not only helps to improve the safety and reliability of the system but also provides a solid data foundation for the future development of autonomous driving technology.

[0106] Step S332: Conduct a quantitative level evaluation on the scoring results of each dimension to obtain the evaluation results of each dimension.

[0107] It should be noted that the quantitative level evaluation means using four levels of G (satisfied), A (good), M (average), and P (dissatisfied) to score the simulation test results of each dimension, providing objective and comparable evaluation results.

[0108] In the evaluation system module, the specific quantification process is to collect relevant simulation test data (such as decision - making time, driving path, interaction response, etc.) from the simulation platform and real - road tests.

[0109] And based on the results of each evaluation dimension obtained in step A21, the following formula is used for quantitative calculation to obtain the evaluation score:

[0110]

[0111] In the formula, N is the number of indicators, and performanceMenteics is the set of performance indicators.

[0112] Among them, the set of performance indicators provides the specific quantitative data required for calculating the comprehensive score. The five - dimensional evaluation system serves as a framework for guiding the selection of the most critical performance indicators and determining how to organize these indicators for a comprehensive evaluation. By using the specific data in the set of performance indicators, this framework can be effectively applied, ensuring a comprehensive and accurate evaluation of the autonomous driving system.

[0113] Finally, according to the calculated evaluation score, map it to one of the four levels of G (satisfied), A (good), M (average), and P (dissatisfied).

[0114] Among them, level G means that the autonomous driving system performs excellently, with almost no need for manual takeover and can handle various complex driving scenarios. At this level, the system can handle sudden situations such as pedestrians suddenly crossing the road or changes in traffic signals with ease. The driver can fully relax and trust the system's judgment, and the vehicle drives smoothly, providing a high - quality riding experience for passengers.

[0115] Level A means that although the autonomous driving system may occasionally produce manual results, its overall performance is still better than average. In most cases, the system can complete driving tasks independently with fewer takeover times. Even when driver assistance is sometimes required in complex scenarios, the overall driving experience is still satisfactory. The system has strong reaction speed and decision-making ability, ensuring driving safety and efficiency.

[0116] Level M means that the autonomous driving system performs mediocrely in some cases and has a high frequency of manual takeover, which may pose some safety hazards. The driver needs to stay alert and be ready to take over control at any time. This uncertainty reduces the passenger's comfort and trust in the system, affecting the overall driving experience.

[0117] Level P means that the autonomous driving system performs poorly and the driver needs to take over frequently, resulting in a lack of trust. The autonomous driving system at this level cannot effectively handle common driving scenarios and has an extremely high number of manual takeover times. The system at this level is difficult to effectively handle common driving scenarios and has an extremely high number of manual takeover times. The driver will feel uneasy and must always concentrate, while passengers may also be skeptical about intelligent driving technology and have a poor experience.

[0118] By adopting a quantitative evaluation level with four dimensions for the simulation test results, the performance and user experience of the intelligent driving system can be more comprehensively understood. This fine-grained evaluation method helps developers identify the deficiencies of the system, so as to make targeted improvements and improve the safety and reliability of intelligent driving.

[0119] Based on the first embodiment of this application, in the third embodiment of this application, the same or similar content as the above-mentioned Embodiment 1 can be referred to the above introduction and will not be repeated hereinafter. On this basis, please refer to Figure 3 , after step S30, the autonomous driving performance optimization method further includes steps S301 to S302:

[0120] Step S301, based on the evaluation result of the autonomous driving performance, optimize the autonomous driving algorithm;

[0121] Specifically, based on the evaluation results of various dimensions obtained in step S30 above, the autonomous driving algorithm is optimized in a targeted manner, and the optimized algorithm is applied to update the simulation platform. For example, the weighted comprehensive score and detailed evaluation report obtained from the simulation test are reviewed to determine the scores of the five evaluation dimensions (such as planning ability, game ability, vehicle control proficiency, environmental perception ability, and human-computer interaction experience). Then, priorities are set for improvement tasks based on the degree of impact of each dimension and the difficulty of technical implementation. For example, if environmental perception ability is a key issue, then it should be put first. And according to the priority of each evaluation dimension, a corresponding optimization strategy is formulated, and the autonomous driving algorithm is rewritten based on the optimization strategy.

[0122] Step S302: Using the optimized autonomous driving algorithm to update the simulation platform.

[0123] Specifically, after obtaining the optimized autonomous driving algorithm, an optimization algorithm (such as gradient descent, genetic algorithm, etc.) is used to adjust the parameters of the simulation platform, and the simulation platform is updated based on the adjusted simulation platform parameters.

[0124] Further, in a feasible embodiment, step S302 may also include step A1:

[0125] Step A1, using the optimized autonomous driving algorithm, adjust the parameters of the simulation platform according to the error between the simulation data and the actual road test results, and update the simulation platform based on the adjusted parameters.

[0126] Specifically, first, the autonomous driving system is run on the simulation platform to collect simulation data such as vehicle status, sensor readings, and environmental parameters. At the same time, in real road tests, actual vehicle operation data is collected, including vehicle trajectory, sensor data, environmental conditions, and other actual vehicle operation data, and the collected data is tested in real road tests based on the optimized autonomous driving algorithm to obtain real road test results.

[0127] Then, the data is preprocessed, including processing missing values, outliers and noise data to ensure data quality; the simulated data and real data are converted into a unified format and unit; the data is timestamped to ensure data alignment in the time dimension.

[0128] Next, the data is fused across domains, such as identifying corresponding features in the simulation data and real data, such as vehicle speed, acceleration, direction, etc. The coordinate system in the simulation environment is aligned with the geographic coordinate system in the real world to ensure the consistency of spatial dimensions. In addition, the same or similar events (such as sudden braking, turning, lane changing, etc.) that occur in the simulation and real data are identified and matched.

[0129] Finally, calculate the error between the simulation data output by the simulation platform and the real road test results, and determine whether the error exceeds a preset threshold. If so, use optimization algorithms (such as gradient descent, genetic algorithms, etc.) to adjust the parameters in the simulation platform according to the error. And update the simulation platform according to the adjusted parameters. The updated simulation platform can output data closer to the real data. Run the updated autonomous driving algorithm again in the real road test environment to generate new real road test results. Then compare whether the error value between the simulation test results and the real road test results exceeds the preset threshold. If it still exceeds, iterate to optimize the autonomous driving algorithm. After each iteration, run the optimized autonomous driving system on the simulation platform again, continue to collect new simulation data, and conduct real road tests again until the difference between the simulation test results and the real road test results falls within the preset threshold.

[0130] Suppose during real road tests, it is found that the braking distance of the vehicle on a wet road surface is longer than predicted by the simulation platform. First, collect the data R(t) of the real vehicle braking test on the wet road surface, and the braking data S(t) under corresponding conditions (such as the same initial speed, vehicle load, weather conditions, etc.) extracted from the simulation platform.

[0131] Then, by calculating the difference between the two data sets E(t) = R(t) - S(t), the deviation between the predicted value of the simulation platform and the actual test results can be quantified. Use algorithms such as the Kalman filter, based on the calculated error E(t), to dynamically adjust the key parameters in the simulation platform - such as the road surface friction coefficient μ. This process aims to minimize the difference between the simulation results and the measured data, enabling the simulation platform to more accurately mimic the real situation.

[0132] Next, update the simulation platform according to the adjusted key parameter values of the real model simulation, especially the updated road surface friction coefficient μ. Then, re-run the simulation of the braking process on the updated model to ensure that the new model settings can accurately reproduce the braking behavior on the wet road surface.

[0133] Finally, apply the improved simulation platform to more wet road surface scenarios to test the braking strategy of the autonomous driving system. If new problems are found, continue data assimilation and model update.

[0134] Through multiple iterations and feedback mechanisms of the above steps, the performance of the simulation platform gradually approaches the real road environment. And, by combining cross-domain data fusion and data assimilation technologies, integrate real road test data into the simulation environment, continuously update and optimize the simulation platform, and improve its accuracy and reliability.

[0135] Through the method of the above embodiments, by optimizing the autonomous driving algorithm to update the simulation platform, the simulation test results can be made closer to the actual situation, which can improve the performance and reliability of the autonomous driving system in different scenarios. Through the process of closed-loop feedback and iterative optimization, the performance of the autonomous driving algorithm can be gradually improved, enabling it to not only work well under ideal simulation conditions but also maintain stable and reliable performance in the face of the uncertainties of the real world.

[0136] Exemplarily, to facilitate understanding of the implementation process of the autonomous driving performance optimization method obtained by combining the above Embodiment 1 to Embodiment 3, please refer to Figure 4 , Figure 4 A brief flowchart of an autonomous driving performance optimization method is provided. Specifically:

[0137] The embodiments of this application include a simulation platform module, an evaluation system module, a real-road evaluation module, a result comparison and optimization module, and a feedback and iteration module. First, input driving scenario parameters, autonomous driving algorithms, etc. into the simulation platform module. During the simulation process of the data on the simulation platform, introduce the evaluation system module, including a five-dimensional evaluation system and a quantitative evaluation level. Score the simulation test results based on the five-dimensional evaluation system and the quantitative evaluation level, including the evaluation of the planning ability, game ability, vehicle control proficiency, environmental perception ability, and human-machine interaction experience of the simulation test results, and conduct a more detailed evaluation of the simulation test results in four dimensions, including satisfactory, good, average, and unsatisfactory. Subsequently, the simulation platform outputs the simulation test results including each evaluation dimension.

[0138] Then, according to the simulation test results, optimize the autonomous driving algorithm and input the optimized autonomous driving algorithm into the real-road evaluation module to verify the optimized autonomous driving algorithm in the actual driving environment, collect real-time data, and obtain the real-road test results.

[0139] Subsequently, input the real-road test results and the simulation test results including each evaluation dimension into the real-road test module for comparison, analyze the performance differences between the two, and propose optimization suggestions to obtain an optimization plan and updated simulation scenario parameters.

[0140] Finally, adjust the simulation test scenario in the feedback and iteration module according to the feedback results of the real-road test, continuously optimize the autonomous driving algorithm, and obtain updated test parameters and an optimized algorithm version. Update the simulation platform based on the updated test parameters and the optimized algorithm version, and repeat the data assimilation and model update process to continuously improve the accuracy of the simulation platform.

[0141] In summary, by combining comprehensive evaluation and feedback mechanisms, the reliability and efficiency of evaluation have been significantly improved. The authenticity of the simulation environment can be further enhanced by introducing more real-world variables and diverse driving behaviors, making the simulation test results closer to the actual situation. In addition, more advanced machine learning algorithms and data analysis techniques can be considered to enhance the adaptability to complex scenarios and the degree of automation of evaluation.

[0142] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the method for optimizing the autonomous driving performance of this application. Based on this technical concept, more forms of simple transformations are within the protection scope of this application.

[0143] This application also provides an autonomous driving performance optimization system. Please refer to Figure 4 , the autonomous driving performance optimization system includes:

[0144] An acquisition module 10, configured to acquire driving scenario parameters and autonomous driving algorithms

[0145] A simulation module 20, configured to construct a simulation platform by using the driving scenario parameters and the autonomous driving algorithms, and perform a simulation test on the autonomous driving system through the simulation platform to obtain a simulation test result;

[0146] An evaluation module 30, configured to adjust the weights and evaluation focuses of the simulation test results through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of the autonomous driving performance.

[0147] The autonomous driving performance optimization system provided by this application adopts the autonomous driving performance optimization method in the above embodiment, and can solve the technical problem of how to improve the reliability and safety of the autonomous driving system in a complex and changeable real driving environment. Compared with the prior art, the beneficial effects of the autonomous driving performance optimization system provided by this application are the same as those of the autonomous driving performance optimization method provided by the above embodiment, and other technical features in the autonomous driving performance optimization system are the same as those disclosed in the method of the above embodiment, and will not be elaborated here.

[0148] This application provides an autonomous driving performance optimization device. The autonomous driving performance optimization device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the autonomous driving performance optimization method in the first embodiment above.

[0149] Next, refer to Figure 6, which shows a schematic structural diagram of an autonomous driving performance optimization device suitable for implementing the embodiments of the present application. The autonomous driving performance optimization device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), in-vehicle terminals (such as in-vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The shown autonomous driving performance optimization device is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0150] As Figure 6 shown, the autonomous driving performance optimization device may include a processing system 1001 (such as a central processing unit, a graphics processing unit, etc.), which may perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage system 1003 into a random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the autonomous driving performance optimization device are also stored. The processing system 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input system 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output system 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage system 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication system 1009. The communication system 1009 may allow the autonomous driving performance optimization device to communicate with other devices wirelessly or wiredly to exchange data. Although the figure shows an autonomous driving performance optimization device having various systems, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be alternatively implemented or had.

[0151] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product that includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through a communication system, or installed from a storage system 1003, or installed from a ROM 1002. When the computer program is executed by a processing system 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are executed.

[0152] The automatic driving performance optimization device provided by the present application adopts the automatic driving performance optimization method in the above embodiment, and can solve the technical problem of how to improve the reliability and safety of the automatic driving system in a complex and changeable real driving environment. Compared with the prior art, the beneficial effects of the automatic driving performance optimization device provided by the present application are the same as those of the automatic driving performance optimization method provided in the above embodiment, and other technical features in the automatic driving performance optimization device are the same as the features disclosed in the method of the previous embodiment, which will not be elaborated here.

[0153] It should be understood that each part disclosed in the present application can be implemented by hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0154] As described above, the above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0155] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the automatic driving performance optimization method in the above embodiment.

[0156] The computer-readable storage medium provided by the present application may, for example, be a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, device, or component. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0157] The above computer-readable storage medium may be included in the autonomous driving performance optimization device; or it may exist separately and not be assembled into the autonomous driving performance optimization device.

[0158] The above computer-readable storage medium carries one or more programs, and when the one or more programs are executed by the autonomous driving performance optimization device, the autonomous driving performance optimization device is caused to: obtain driving scenario parameters and an autonomous driving algorithm; use the driving scenario parameters and the autonomous driving algorithm to construct a simulation platform, and perform a simulation test on the autonomous driving system through the simulation platform to obtain a simulation test result; adjust the weights and evaluation focuses of the simulation test result through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of the autonomous driving performance.

[0159] Computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The above-mentioned programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, by connecting through the Internet using an Internet service provider).

[0160] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and this module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutively represented blocks can actually be executed substantially in parallel, and they can sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0161] The modules involved in the embodiments described in this application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation on the unit itself in some cases.

[0162] The readable storage medium provided in this application is a computer-readable storage medium. The computer-readable storage medium stores computer-readable program instructions (i.e., computer programs) for performing the above-mentioned autonomous driving performance optimization method, and can solve the technical problems of autonomous driving performance optimization. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the autonomous driving performance optimization method provided in the above embodiments, and will not be elaborated here.

[0163] The present application also provides a computer program product, including a computer program which, when executed by a processor, implements the steps of the above-mentioned method for optimizing autonomous driving performance.

[0164] The computer program product provided by the present application can solve the technical problem of how to improve the reliability and safety of an autonomous driving system in a complex and changeable real driving environment. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as those of the method for optimizing autonomous driving performance provided in the above embodiments, and will not be elaborated herein.

[0165] The foregoing are only partial embodiments of the present application, and thus do not limit the patent scope of the present application. Any equivalent structural transformation made by using the content of the specification and drawings of the present application under the technical concept of the present application, or any direct / indirect application in other related technical fields shall be included within the patent protection scope of the present application.

Claims

1. A method for optimizing the performance of an automatic driving system, characterized in that: The method comprises: Obtain driving scenario parameters and autonomous driving algorithms; Using the driving scenario parameters and the autonomous driving algorithm to build a simulation platform, and using the simulation platform to perform a simulation test on the autonomous driving system to obtain a simulation test result; The weights and evaluation focus of the simulation test results are adjusted through a pre-constructed multi-dimensional evaluation system to obtain an evaluation result of the autonomous driving performance.

2. The method according to claim 1, characterized in that The step of adjusting the weight and evaluation focus of the simulation test results through the pre-built multi-dimensional evaluation system to obtain the evaluation results of the autonomous driving performance includes: Extracting and analyzing the driving scene parameters to obtain a scene feature vector; Calculate and adjust the weight of the multi-dimensional evaluation system according to the scene feature vector; According to the adjusted weights of the multi-dimensional evaluation system and the evaluation results of each evaluation dimension, the evaluation focus of the multi-dimensional evaluation system is determined, and a weighted comprehensive score is generated; Based on the weighted comprehensive score, a weighted comprehensive evaluation is performed on the simulation test results to obtain an evaluation result of the autonomous driving performance.

3. The method according to claim 2, characterized in that Before the step of confirming the evaluation focus of the multi-dimensional evaluation system and generating a weighted comprehensive score based on the adjusted weights of the multi-dimensional evaluation system and the evaluation results of each evaluation dimension, the step includes: The simulation test results are scored and calculated through a pre-built multi-dimensional evaluation system to obtain the scoring results of each dimension; The scoring results of the various dimensions are quantitatively evaluated to obtain evaluation results of the various evaluation dimensions.

4. The method according to claim 3, characterized in that The pre-built multi-dimensional evaluation system at least includes planning ability evaluation, game ability evaluation, and vehicle control proficiency evaluation; The step of scoring the simulation test results by using the pre-built multi-dimensional evaluation system to obtain the scoring results of each dimension includes: For the planning capability assessment, calculating the deviation between the planned path and the optimal path in the simulation test result, and the ratio between the travel time and the shortest travel time, and calculating the planning capability score based on the deviation and the ratio; and / or For the gaming ability evaluation, identifying the set of interactive events in the simulation test results, determining whether the decision of each event in the interactive event set complies with preset rules, obtaining a decision score, and calculating the gaming ability score based on the decision score; and / or For the vehicle control proficiency assessment, the variance of the vehicle acceleration change rate and the mean square error of the vehicle deviation from the expected trajectory in the simulation test results are calculated, and the vehicle control proficiency score is calculated based on the variance and the mean square error.

5. The method according to claim 3, characterized in that The step of performing quantitative level evaluation on the initial evaluation result to obtain evaluation results of each evaluation dimension includes: Performing quantitative calculation on the initial evaluation result to obtain an evaluation score; The evaluation scores are mapped to preset quantitative levels to obtain evaluation results of various evaluation dimensions.

6. The method according to claim 1, characterized in that After the step of adjusting the weight and evaluation focus of the simulation test results through the pre-built multi-dimensional evaluation system to obtain the evaluation results of the autonomous driving performance, the method further includes: Optimizing the autonomous driving algorithm based on the evaluation result of the autonomous driving performance; The optimized autonomous driving algorithm is used to update the simulation platform.

7. The method according to claim 6, characterized in that The step of using the optimized autonomous driving algorithm to update the simulation platform includes: Utilizing the optimized autonomous driving algorithm, the parameters of the simulation platform are adjusted according to the error between the simulation data and the actual road test results, and the simulation platform is updated based on the adjusted parameters.

8. The method according to claim 7, characterized in that The step of using the optimized autonomous driving algorithm to adjust the parameters of the simulation platform according to the error between the simulation data and the actual road test results includes: Obtain real road test results based on the optimized autonomous driving algorithm; Comparing whether the error between the simulation test result and the real road test result is greater than a preset threshold; If so, the parameters of the simulation platform are adjusted according to the error between the simulation test result and the real road test result until the difference between the simulation test result and the real road test result is less than the preset threshold.

9. An automatic driving performance optimization system, characterized in that: The system comprises: Acquisition module, used to obtain driving scene parameters and autonomous driving algorithms; A simulation module, used to build a simulation platform using the driving scenario parameters and the autonomous driving algorithm, and to perform a simulation test on the autonomous driving system through the simulation platform to obtain a simulation test result; The evaluation module is used to adjust the weight and evaluation focus of the simulation test results through a pre-built multi-dimensional evaluation system to obtain the evaluation results of the autonomous driving performance.

10. An automatic driving performance optimization device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the method for optimizing the performance of an autonomous driving as described in any one of claims 1 to 8.

11. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the automatic driving performance optimization method as described in any one of claims 1 to 8 are implemented.

12. A computer program product, characterized in that The computer program product comprises a computer program, which, when executed by a processor, implements the steps of the method for optimizing the performance of an autonomous driving as claimed in any one of claims 1 to 8.

Images