Methods for Improving Autonomous Driving Test Cases Based on Orthogonal Experiments

Abstract

This invention discloses a method for improving autonomous driving test cases based on orthogonal experiments, comprising: determining the factors of the orthogonal experiment based on the scenario to be analyzed; determining the level corresponding to each factor of the orthogonal experiment; determining the experimental evaluation strategy of the orthogonal experiment; designing autonomous driving test cases based on the factors of the orthogonal experiment and the corresponding levels of each factor; conducting tests based on the autonomous driving test cases; evaluating the test results of the test cases according to the experimental evaluation strategy; filtering the levels corresponding to each factor based on the contribution rate; and updating the test cases based on the contribution rate filtering results. The method for improving autonomous driving test cases based on orthogonal experiments provided by this invention filters out valuable parameters in the test cases by designing orthogonal experiments, and then increases the sampling number near these values ​​to improve the typicality of the autonomous driving test cases, deletes low-value test cases, and increases the value of the test cases, thereby improving testing efficiency while ensuring test coverage.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and in particular to a method for improving autonomous driving test cases based on orthogonal experiments. Background Technology

[0002] Currently, designing test cases based on driving scenarios to test and improve existing autonomous driving systems has become the mainstream approach for testing L2 and higher autonomous driving systems. This involves deconstructing driving scenarios to generate elements that constitute the scenario, statistically analyzing the probability of each element appearing in the samples, and then combining these probabilities into test cases. However, a significant problem arises: increasing the number of possible values ​​for scenario elements leads to a sharp increase in the number of generated test cases. Since the generated test cases are often homogeneous and contain many common low-value scenarios, this results in inefficiency during subsequent testing.

[0003] Therefore, there is an urgent need for a method to improve autonomous driving test cases based on orthogonal experiments. Summary of the Invention

[0004] The purpose of this invention is to provide a method for improving autonomous driving test cases based on orthogonal experiments, in order to solve the problems in the prior art, improve the value of test cases, and increase test efficiency while ensuring test coverage.

[0005] This invention provides a method for improving autonomous driving test cases based on orthogonal experiments, comprising the following steps:

[0006] Determine the factors for orthogonal experiments based on the scenario to be analyzed;

[0007] Determine the levels corresponding to each factor in the orthogonal experiment;

[0008] Determine the experimental evaluation strategy for orthogonal experiments;

[0009] Based on the factors of the orthogonal experiment and the corresponding levels of each factor, design autonomous driving test cases and conduct tests based on the autonomous driving test cases;

[0010] The test results of the autonomous driving test cases are evaluated according to the test evaluation strategy, and the levels corresponding to each factor are screened based on the contribution rate.

[0011] Based on the screening results of contribution rate, the autonomous driving test cases are updated.

[0012] The method for improving autonomous driving test cases based on orthogonal experiments, as described above, preferably includes the following step: determining the factors for orthogonal experiments based on the scenario to be analyzed.

[0013] The collected natural driving data of the same type were deconstructed;

[0014] At the same time, based on the scenario to be tested, scenario parameters that have a significant impact on the scenario are selected as factors for orthogonal experimental design.

[0015] The method for improving autonomous driving test cases based on orthogonal experiments, as described above, preferably includes the following step: Deconstructing the collected natural driving data of the same type, specifically comprising:

[0016] The collected data is deconstructed based on a seven-layer scenario library architecture, which includes: roads, traffic facilities, temporary changes to roads and facilities, traffic participants, climate environment, communication information, and vehicle status.

[0017] The method for improving autonomous driving test cases based on orthogonal experiments, as described above, preferably includes, in determining the levels corresponding to each factor in the orthogonal experiments, the following specific steps:

[0018] Determine the variable type for each factor;

[0019] If the variable type of the factor is an interval size variable, then determining the several levels corresponding to each factor in the orthogonal experiment specifically includes:

[0020] The sample size of each factor is counted. According to the value of the factor corresponding to each sample from smallest to largest, the sample of the factor is divided into multiple groups on average. The number of groups is the number of levels. The average value of the factor in each group is taken as the level of the factor.

[0021] For the nominal size variable, the determination of the levels corresponding to each factor in the orthogonal experiment specifically includes:

[0022] The values ​​of all variables are used as the level of this factor.

[0023] The method for improving autonomous driving test cases based on orthogonal experiments as described above, preferably, includes the following: Determining the test evaluation strategy for orthogonal experiments specifically includes:

[0024] Determine multiple test indicators for orthogonal experiments, wherein the multiple test indicators include perception test indicators, decision test indicators, and execution test indicators;

[0025] The reference variables corresponding to each of the aforementioned test indicators are determined, wherein the reference variables corresponding to the perception test indicators include accuracy, perception speed, comprehensiveness, and stability; the reference variables corresponding to the decision test indicators include decision cycle and risk level; and the reference variables corresponding to the execution test indicators include acceleration, swiftness, and reaction speed.

[0026] Determine the evaluation score of each test indicator, and perform a weighted summation of the evaluation scores of the reference variables corresponding to each test indicator to obtain the evaluation score of the test indicator corresponding to that test indicator.

[0027] The comprehensive test score is obtained by weighted summation of the test index evaluation scores corresponding to each of the aforementioned test indicators.

[0028] The method for improving autonomous driving test cases based on orthogonal experiments, as described above, preferably includes the following steps: designing autonomous driving test cases based on the factors of the orthogonal experiment and the corresponding levels of each factor, and conducting tests based on the autonomous driving test cases.

[0029] Design test cases for autonomous driving based on the design rules, factors, and levels of each factor in orthogonal experiments.

[0030] Use simulation software to draw the aforementioned autonomous driving test cases;

[0031] In each scenario to be analyzed, the aforementioned autonomous driving test cases will be tested according to perception test indicators, and / or decision test indicators, and / or execution test indicators.

[0032] The method for improving autonomous driving test cases based on orthogonal experiments as described above, preferably, involves testing each autonomous driving test case according to perception test indicators, and / or decision test indicators, and / or execution test indicators in each scenario to be analyzed, specifically including:

[0033] For the sensing test indicators, the sensor hardware-in-the-loop test method is used to test each sensor;

[0034] For decision-making test indicators, the controller hardware-in-the-loop test method is used to test each control algorithm;

[0035] To implement the test targets, the vehicle-in-the-loop test method is used to test the chassis of the actual vehicle.

[0036] The method for improving autonomous driving test cases based on orthogonal experiments as described above, preferably, further includes testing each of the autonomous driving test cases according to perception test indicators, and / or decision test indicators, and / or execution test indicators in each scenario to be analyzed, and further includes:

[0037] During the testing process, the test indicators corresponding to the same autonomous driving test case are repeated multiple times, and the average value of the multiple test results is taken as the test result of the autonomous driving test case.

[0038] The method for improving autonomous driving test cases based on orthogonal experiments as described above, preferably, involves evaluating the test results of the autonomous driving test cases according to an experimental evaluation strategy, and screening the levels corresponding to each factor based on the contribution rate, specifically including:

[0039] For each factor in the orthogonal experiment, the level with the highest evaluation score of the experimental index is taken as the level with the largest contribution rate under that factor, and the level with the lowest evaluation score of the experimental index is taken as the level with the smallest contribution rate under that factor.

[0040] The method for improving autonomous driving test cases based on orthogonal experiments as described above, preferably, involves updating the autonomous driving test cases based on the screening results of contribution rates, specifically including:

[0041] For each factor in the orthogonal experiment, the autonomous driving test cases corresponding to the level values ​​with small contribution rates were deleted, and the sampling values ​​around the level values ​​with large contribution rates were added.

[0042] The present invention provides a method for improving autonomous driving test cases based on orthogonal experiments. By designing orthogonal experiments on the basis of existing experiments, valuable parameters in the test cases are screened out. Then, the sampling number is increased near these values ​​to improve the typicality of autonomous driving test cases, delete low-value test cases, and increase the value of test cases. This improves test efficiency while ensuring test coverage. Attached Figure Description

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described below with reference to the accompanying drawings, wherein:

[0044] Figure 1 A flowchart illustrating an embodiment of the method for improving autonomous driving test cases based on orthogonal experiments provided by the present invention;

[0045] Figure 2 This is a logic diagram of an embodiment of the method for improving autonomous driving test cases based on orthogonal experiments provided by the present invention. Detailed Implementation

[0046] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. The descriptions of the exemplary embodiments are merely illustrative and are in no way intended to limit the present disclosure or its application or use. The present disclosure may be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided so that the present disclosure will be thorough and complete, and will fully express the scope of the disclosure to those skilled in the art. It should be noted that, unless specifically stated otherwise, the relative arrangement of components and steps, the composition of materials, numerical expressions, and values ​​set forth in these embodiments should be interpreted as exemplary only and not as limiting.

[0047] The terms “first,” “second,” and similar terms used in this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different parts. Terms such as “including” or “contains” mean that the element preceding the term encompasses the element listed after it, and do not exclude the possibility of encompassing other elements as well. Terms such as “above” and “below” are used only to indicate relative positional relationships; when the absolute position of the described object changes, this relative positional relationship may also change accordingly.

[0048] In this disclosure, when a specific component is described as being located between a first component and a second component, an intermediary component may or may not be present between the specific component and the first or second component. When a specific component is described as connecting to other components, the specific component may be directly connected to the other components without having an intermediary component, or it may not be directly connected to the other components but may have an intermediary component.

[0049] All terms used in this disclosure (including technical or scientific terms) have the same meaning as understood by one of ordinary skill in the art to which this disclosure pertains, unless otherwise specifically defined. It should also be understood that terms defined in a general dictionary, such as a dictionary, should be interpreted as having a meaning consistent with their meaning in the context of the relevant art, and not as having an idealized or highly formalized meaning, unless expressly defined herein.

[0050] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, they should be considered part of the specification.

[0051] As mentioned earlier, in current scenario-based testing of autonomous driving systems, the numerous possible values ​​for scenario elements after scenario deconstruction lead to a sharp increase in the number of test cases generated. These test cases include both high-value and low-value ones, while low-value or homogeneous scenarios contribute little to the testing of autonomous driving systems. Therefore, this invention utilizes orthogonal experimental design to design test cases, identifying important parameter values ​​within the test cases. This increases the value of the test cases, thereby covering more critical scenarios. This approach reduces the number of low-value scenario test cases while increasing the number of high-value scenario test cases.

[0052] In terms of current scenario generation for testing autonomous driving systems, there are two main approaches: expert experience and data collected from natural driving. This invention mainly uses statistical analysis based on data collected from natural driving to generate test cases.

[0053] like Figure 1 and Figure 2As shown, the method for improving autonomous driving test cases based on orthogonal experiments provided in this embodiment includes the following steps in actual execution:

[0054] Step S1: Determine the factors for the orthogonal experiment based on the scenario to be analyzed.

[0055] In one embodiment of the method for improving autonomous driving test cases based on orthogonal experiments of the present invention, step S1 may specifically include:

[0056] Step S11: Deconstruct the collected natural driving data of the same type.

[0057] Specifically, the collected data is deconstructed according to a seven-layer scenario library architecture (referencing "Road Vehicles Expected Functional Safety" (IOS / PAS 21448:2019)). This seven-layer scenario library architecture includes: roads, traffic facilities, temporary changes to roads and facilities, traffic participants, climate conditions, communication information, and vehicle status.

[0058] Step S12: Simultaneously, based on the scenario to be tested, select the scenario parameters that have a significant impact on the scenario as factors for orthogonal experimental design.

[0059] Step S2: Determine the levels corresponding to each factor in the orthogonal experiment.

[0060] In one embodiment of the method for improving autonomous driving test cases based on orthogonal experiments of the present invention, step S2 may specifically include:

[0061] Step S21: Determine the variable type of each factor.

[0062] Step S22: If the variable type of the factor is an interval size variable, then determining the number of levels corresponding to each factor in the orthogonal experiment specifically includes: counting the number of samples of each factor collected, dividing the sample of the factor into multiple groups in ascending order of the value of the factor corresponding to each sample, the number of groups being the number of levels, and the average value of the factor in each group being taken as the level of the factor.

[0063] For example, suppose a factor has 6 levels. If n samples are generated for this factor during sampling, these n samples are divided into 6 groups, and the average value of each group is taken. The level of this factor, where n represents the number of samples, x i This represents the value of the i-th sample.

[0064] Step S23: For the nominal size variable, determining the levels corresponding to each factor in the orthogonal experiment specifically includes: taking the values ​​of all variables as the levels of the factor.

[0065] Step S3: Determine the experimental evaluation strategy for orthogonal experiments.

[0066] In one embodiment of the method for improving autonomous driving test cases based on orthogonal experiments of the present invention, step S3 may specifically include:

[0067] Step S31: Determine multiple test indicators for the orthogonal experiment, wherein the multiple test indicators include perception test indicators, decision test indicators, and execution test indicators.

[0068] Step S32: Determine the reference variables corresponding to each of the test indicators, wherein the reference variables corresponding to the perception test indicators include accuracy, perception speed, comprehensiveness and stability; the reference variables corresponding to the decision test indicators include decision cycle and risk level; and the reference variables corresponding to the execution test indicators include acceleration, swiftness and reaction speed.

[0069] Since autonomous driving systems comprise three aspects—perception, decision-making, and execution—the test cases designed in this invention also evaluate these three aspects. Therefore, these three aspects are used as indicators for orthogonal experiments. The perception part includes four aspects: the accuracy of perceiving environmental objects, the speed of perceiving the environment, the comprehensiveness of perceiving the environment, and the stability of perceiving the environment. The decision-making system is evaluated by the decision-making time and risk level (TTC). The execution system is evaluated by the acceleration and jerkiness of execution. The test indicators and corresponding reference variables are shown in Table 1.

[0070] Table 1. Experimental Indicators and Corresponding Reference Variables

[0071]

[0072] Step S33: Determine the evaluation score of each test indicator, and perform a weighted summation of the evaluation scores of the reference variables corresponding to each test indicator to obtain the evaluation score of the test indicator corresponding to that test indicator.

[0073] Since the evaluation result of each experimental indicator is determined by multiple variables, a weight needs to be added to each reference variable value when scoring, and the calculated weight value is used as the experimental result of each experimental indicator. Taking the perception indicator as an example, the scores of accuracy, perception speed, comprehensiveness, and stability are weighted and summed to obtain the experimental indicator evaluation score of the perception indicator.

[0074] Step S34: Weighted summation of the evaluation scores of each test index to obtain the comprehensive test score.

[0075] The comprehensive test score is obtained by weighting and summing the evaluation scores of the test indicators corresponding to the perception indicators, decision indicators, and execution indicators.

[0076] Step S4: Based on the factors of the orthogonal experiment and the corresponding levels of each factor, design autonomous driving test cases and conduct tests based on the autonomous driving test cases.

[0077] After determining the factors and their corresponding levels in the orthogonal experiment, step S4 begins by designing test cases according to the design rules of orthogonal experiments. In one embodiment of the method for improving autonomous driving test cases based on orthogonal experiments of the present invention, step S4 may specifically include:

[0078] Step S41: Design autonomous driving test cases according to the design rules, factors and corresponding levels of each factor in orthogonal experiments.

[0079] Step S42: Use simulation software to draw each of the aforementioned autonomous driving test cases.

[0080] Each test case is plotted using simulation software to prepare for subsequent experiments.

[0081] Step S43: In each scenario to be analyzed, each of the autonomous driving test cases is tested according to perception test indicators, and / or decision test indicators, and / or execution test indicators.

[0082] To control for a single variable, this invention conducts different experiments on a scenario based on three indicators. In some embodiments of this invention, only one indicator may be tested, or two or three indicators may be tested simultaneously; this invention does not specifically limit this. In one embodiment of the method for improving autonomous driving test cases based on orthogonal experiments of this invention, step S43 may specifically include:

[0083] Step S431: For the sensing test indicators, the sensor hardware-in-the-loop test method is used to test each sensor.

[0084] For the testing indicators that are related to perception, it is necessary to conduct tests on the sensors that perceive the environment. Therefore, this invention uses sensor hardware-in-the-loop to test each sensor.

[0085] Step S432: For decision test indicators, the controller hardware-in-the-loop test method is used to test each control algorithm.

[0086] Since the test indicators are part of the decision-making process, the control algorithm needs to be tested. Therefore, this invention uses the hardware-in-the-loop of the controller for testing.

[0087] Step S433: To perform the test indicators, the chassis of the actual vehicle is tested using the whole vehicle in-loop test method.

[0088] For the execution part, it is necessary to test the chassis of the actual vehicle. Therefore, the present invention conducts testing through whole vehicle in-loop testing. That is, the present invention uses three different testing methods to evaluate different test indicators.

[0089] In some embodiments of the present invention, step S43 further includes:

[0090] Step S434: During the testing process, the test indicators corresponding to the same autonomous driving test case are repeatedly tested multiple times, and the average value of the multiple test results is taken as the test result of the autonomous driving test case.

[0091] The same experiment needs to be repeated multiple times, and the average of the results from multiple sets of experiments is taken as the result of that test case metric. When only one test metric from perception, decision-making, and execution needs to be tested, such as the decision metric, the test result of the autonomous driving test case is the average of the test metric evaluation scores corresponding to the decision metric in multiple experiments. When two or three test metrics from perception, decision-making, and execution need to be tested, such as the decision metric and the decision indicator, the test result of the autonomous driving test case is the average of the weighted sum of the test metric evaluation scores corresponding to the decision metric and the decision indicator in multiple experiments.

[0092] Step S5: Evaluate the test results of the autonomous driving test cases according to the test evaluation strategy, and filter the levels corresponding to each factor based on the contribution rate.

[0093] Specifically, for each factor in the orthogonal experiment, the level with the highest evaluation score of the experimental indicator is designated as the level with the largest contribution rate under that factor, while the level with the lowest evaluation score is designated as the level with the smallest current contribution rate under that factor. This invention, by determining the most important level value for each experimental indicator and the level value with the smallest contribution rate in the current test case, can filter out the level values ​​with the largest contribution rate and the level values ​​with the smallest current contribution rate.

[0094] Step S6: Update the autonomous driving test cases based on the screening results of the contribution rate.

[0095] Specifically, for each factor in the orthogonal experiment, the autonomous driving test cases corresponding to the level values ​​with small contribution rates are deleted, and the sampling values ​​around the level values ​​with large contribution rates are added.

[0096] Step S6 removes the level values ​​with low contribution rates selected in step S5 from the test cases, and then increases the sampled values ​​around the level values ​​with high contribution rates to optimize the test cases. Meanwhile, as... Figure 2 As shown, it is also possible to choose to return to step S4 for further iterative optimization based on the optimization.

[0097] In simulation experiments, this invention evaluates the importance of parameters in each test case generated through orthogonal experiments from three dimensions: perception, decision-making, and execution. That is, it evaluates the value of the level value. The parameter values ​​with a greater impact on the result are extracted, and the sampling amount around the value is increased. For parameters with a smaller impact, the sampling amount around the parameter value is reduced. This improves the value of the test cases and thus improves the efficiency of the test.

[0098] The method for improving autonomous driving test cases based on orthogonal experiments provided in this invention improves the typicality of autonomous driving test cases by designing orthogonal experiments on the basis of the original experiments, and then increases the sampling number near the value to remove low-value test cases and increase the value of test cases, thereby improving test efficiency while ensuring test coverage.

[0099] The embodiments of this disclosure have now been described in detail. To avoid obscuring the concept of this disclosure, some details known in the art have not been described. Those skilled in the art can fully understand how to implement the technical solutions disclosed herein based on the above description.

[0100] While specific embodiments of this disclosure have been described in detail by way of examples, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of this disclosure. Those skilled in the art should understand that modifications can be made to the above embodiments or equivalent substitutions can be made to some technical features without departing from the scope and spirit of this disclosure. The scope of this disclosure is defined by the appended claims.

Claims

1. A method for improving autonomous driving test cases based on orthogonal experiments, characterized in that, Includes the following steps: Determine the factors for orthogonal experiments based on the scenario to be analyzed; Determine the levels corresponding to each factor in the orthogonal experiment; Determine the experimental evaluation strategy for orthogonal experiments; Based on the factors of the orthogonal experiment and the corresponding levels of each factor, design autonomous driving test cases and conduct tests based on the autonomous driving test cases; The test results of the autonomous driving test cases are evaluated according to the test evaluation strategy, and the levels corresponding to each factor are screened based on the contribution rate. Based on the screening results of contribution rate, the autonomous driving test cases are updated. The experimental evaluation strategy for determining orthogonal experiments specifically includes: Determine multiple test indicators for orthogonal experiments, wherein the multiple test indicators include perception test indicators, decision test indicators, and execution test indicators; The reference variables corresponding to each of the aforementioned test indicators are determined, wherein the reference variables corresponding to the perception test indicators include accuracy, perception speed, comprehensiveness, and stability; the reference variables corresponding to the decision test indicators include decision cycle and risk level; and the reference variables corresponding to the execution test indicators include acceleration, swiftness, and reaction speed. Determine the evaluation score of each test indicator, and perform a weighted summation of the evaluation scores of the reference variables corresponding to each test indicator to obtain the evaluation score of the test indicator corresponding to that test indicator. The comprehensive test score is obtained by weighted summing of the evaluation scores corresponding to each of the aforementioned test indicators. The evaluation of the test results of the autonomous driving test cases according to the test evaluation strategy, and the screening of each level corresponding to each factor based on the contribution rate, specifically includes: For each factor in the orthogonal experiment, the level with the highest evaluation score of the experimental index is taken as the level with the largest contribution rate under that factor, and the level with the lowest evaluation score of the experimental index is taken as the level with the smallest contribution rate under that factor.

2. The method for improving autonomous driving test cases based on orthogonal experiments according to claim 1, characterized in that, The factors for determining the orthogonal experiment based on the scenario to be analyzed specifically include: The collected natural driving data of the same type were deconstructed; At the same time, based on the scenario to be tested, scenario parameters that have a significant impact on the scenario are selected as factors for orthogonal experimental design.

3. The method for improving autonomous driving test cases based on orthogonal experiments according to claim 2, characterized in that, The process of deconstructing the collected natural driving data of the same type specifically includes: The collected data is deconstructed based on a seven-layer scenario library architecture, which includes: roads, traffic facilities, temporary changes to roads and facilities, traffic participants, climate environment, communication information, and vehicle status.

4. The method for improving autonomous driving test cases based on orthogonal experiments according to claim 1, characterized in that, The determination of the levels corresponding to each factor in the orthogonal experiment specifically includes: Determine the variable type for each factor; If the variable type of the factor is an interval size variable, then determining the several levels corresponding to each factor in the orthogonal experiment specifically includes: The sample size of each factor is counted. According to the value of the factor corresponding to each sample from smallest to largest, the sample of the factor is divided into multiple groups on average. The number of groups is the number of levels. The average value of the factor in each group is taken as the level of the factor. For the nominal size variable, the determination of the levels corresponding to each factor in the orthogonal experiment specifically includes: The values ​​of all variables are used as the level of this factor.

5. The method for improving autonomous driving test cases based on orthogonal experiments according to claim 1, characterized in that, The process involves designing autonomous driving test cases based on the factors of the orthogonal experiment and the corresponding levels of each factor, and then conducting tests based on these test cases. Specifically, this includes: Design test cases for autonomous driving based on the design rules, factors, and levels of each factor in orthogonal experiments. Use simulation software to draw the aforementioned autonomous driving test cases; In each scenario to be analyzed, the aforementioned autonomous driving test cases will be tested according to perception test indicators, and / or decision test indicators, and / or execution test indicators.

6. The method for improving autonomous driving test cases based on orthogonal experiments according to claim 5, characterized in that, In each scenario to be analyzed, each of the autonomous driving test cases will be tested according to perception test indicators, and / or decision test indicators, and / or execution test indicators, specifically including: For the sensing test indicators, the sensor hardware-in-the-loop test method is used to test each sensor; For decision-making test indicators, the controller hardware-in-the-loop test method is used to test each control algorithm; To implement the test targets, the vehicle-in-the-loop test method is used to test the chassis of the actual vehicle.

7. The method for improving autonomous driving test cases based on orthogonal experiments according to claim 6, characterized in that, The step of testing each autonomous driving test case in each scenario to be analyzed according to perception test indicators, and / or decision test indicators, and / or execution test indicators, further includes: During the testing process, the test indicators corresponding to the same autonomous driving test case are repeated multiple times, and the average value of the multiple test results is taken as the test result of the autonomous driving test case.

8. The method for improving autonomous driving test cases based on orthogonal experiments according to claim 1, characterized in that, The process of updating autonomous driving test cases based on the screening results of contribution rate specifically includes: For each factor in the orthogonal experiment, the autonomous driving test cases corresponding to the level values ​​with small contribution rates were deleted, and the sampling values ​​around the level values ​​with large contribution rates were added.

Images