An automatic driving simulation test method, device, equipment and storage medium

Abstract

The application discloses an automatic driving simulation test method and device, equipment and a storage medium. Complete road section data is cut to obtain a plurality of data segments; each frame of data in the data segments is input into a first benchmark path planning algorithm and a to-be-compared path planning algorithm to perform path planning respectively, path planning trajectories corresponding to a preset future time period are obtained, and whether there is a difference frame in the data segments is determined by comparing the path planning trajectories; the difference frames in a target data segment are grouped to generate a plurality of difference scenes; in the difference scenes, the first benchmark path planning algorithm and the to-be-compared path planning algorithm are used for simulation test respectively, corresponding simulation results are obtained, and the to-be-compared path planning algorithm is evaluated according to the simulation results corresponding to the first benchmark path planning algorithm and the to-be-compared path planning algorithm, thereby improving the technical problem that the automatic driving capability changes caused by the fact that the prior art cannot test path planning algorithm iteration in simulation test.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, and in particular to an automatic driving simulation test method, device, equipment and storage medium. BACKGROUND

[0002] The existing simulation test method is to collect data of abnormal scenes in real road tests, and then reproduce the scenes in a virtual environment for simulation testing. This simulation test method has some limitations, for example, in a normally running road section, if a new accident is caused by path planning algorithm, the existing simulation test method only collects scenes that appear abnormally in real road tests, does not collect data of this section, and thus cannot obtain this kind of problem, and cannot test the changes of automatic driving capability caused by path planning algorithm iteration in simulation testing. SUMMARY

[0003] The present application provides an automatic driving simulation test method, device, equipment and storage medium, which is used to solve the technical problem that the existing technology cannot test the changes of automatic driving capability caused by path planning algorithm iteration in simulation testing.

[0004] Therefore, the first aspect of the present application provides an automatic driving simulation test method, comprising:

[0005] segmenting the collected complete road section data to obtain a plurality of data segments;

[0006] inputting each frame of data in the data segment into a first benchmark path planning algorithm and a to-be-compared path planning algorithm for path planning respectively to obtain path planning trajectories corresponding to a preset future time period, and determining whether there is a difference frame in the data segment according to the path planning trajectories corresponding to the first benchmark path planning algorithm and the to-be-compared path planning algorithm to obtain a target data segment with a difference frame;

[0007] grouping the difference frames in the target data segment to generate a plurality of difference scenes;

[0008] in the difference scenes, performing simulation testing by using the first benchmark path planning algorithm and the to-be-compared path planning algorithm respectively to obtain corresponding simulation results, and evaluating the to-be-compared path planning algorithm according to the simulation results corresponding to the first benchmark path planning algorithm and the to-be-compared path planning algorithm.

[0009] Optionally, the determination of whether there is a difference frame in the data segment according to the path planning trajectories corresponding to the first benchmark path planning algorithm and the to-be-compared path planning algorithm to obtain a target data segment with a difference frame comprises:

[0010] Calculate the difference between the path planning trajectories output by the first baseline path planning algorithm and the path planning algorithm to be compared for the same frame of data;

[0011] When the difference value is greater than a preset distance threshold, the data frame corresponding to the difference value is determined to be a difference frame;

[0012] The data segment containing the differing frame is taken as the target data segment.

[0013] Optionally, the step of grouping the difference frames in the target data segment to generate several difference scenarios includes:

[0014] Sort the difference frames in the target data segment in ascending order by time;

[0015] Merge the difference frames in the target data segment whose time difference is less than a preset time threshold to obtain the difference frame set of the target data segment;

[0016] The set of difference frames of the target data segment is divided into several difference scenes.

[0017] Optionally, in the difference scenario, simulation tests are performed using the first baseline path planning algorithm and the path planning algorithm to be compared, respectively, to obtain corresponding simulation results, including:

[0018] The first frame data in the difference scenario is used as the current frame and input into the first benchmark path planning algorithm and the path planning algorithm to be compared respectively for path planning. The simulation vehicle is controlled to drive according to the path planning trajectory of the future preset time period obtained from the current frame.

[0019] The next frame data in the difference scenario is used as the current frame and input into the first benchmark path planning algorithm and the path planning algorithm to be compared for path planning. The simulation vehicle is controlled to drive according to the path planning trajectory of the future preset time period obtained from the current frame until all data frames in the difference scenario are tested, and the simulation results of the first benchmark path planning algorithm and the path planning algorithm to be compared in the difference scenario are obtained.

[0020] Optionally, evaluating the path planning algorithm to be compared based on the simulation results corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared includes:

[0021] Based on the selected scenario metric, the simulation results of the first baseline path planning algorithm in each of the different scenarios are measured to obtain a first metric value for each of the different scenarios. The scenario metric includes scenario safety and / or scenario comfort.

[0022] Based on the selected scenario metric, the simulation results of the path planning algorithm to be compared in each of the different scenarios are measured to obtain a second metric value for each of the different scenarios;

[0023] The path planning algorithms to be compared are evaluated using the first and second metrics for each of the different scenarios.

[0024] Optionally, evaluating the path planning algorithm to be compared using the first and second metrics for each of the said difference scenarios includes:

[0025] Compare the first metric and the second metric for the same difference scenario;

[0026] When the second metric value of the same difference scenario is greater than the corresponding first metric value, the difference scenario is determined to be an improved scenario;

[0027] When the second metric value of the same difference scenario is less than the corresponding first metric value, the difference scenario is determined to be a variation scenario;

[0028] The path planning algorithm to be compared is evaluated based on the number of improved scenarios and the number of worse scenarios.

[0029] Optionally, the method further includes:

[0030] In the aforementioned difference scenario, a second benchmark path planning algorithm is used for simulation testing, and the evaluation results obtained by the path planning algorithm to be compared are verified based on the simulation results obtained by the second benchmark path planning algorithm.

[0031] Optionally, the number of path planning algorithms to be compared is one or more.

[0032] A second aspect of this application provides an autonomous driving simulation testing device, comprising:

[0033] The segmentation unit is used to segment the collected complete road segment data into several data fragments;

[0034] The path planning unit is used to input each frame of data in the data segment into the first benchmark path planning algorithm and the path planning algorithm to be compared, respectively to perform path planning, obtain the path planning trajectory for the corresponding future preset time period, and determine whether there are difference frames in the data segment based on the path planning trajectories corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared, so as to obtain the target data segment with difference frames.

[0035] The generation unit is used to group the difference frames in the target data segment and generate several difference scenes;

[0036] The simulation unit is used to perform simulation tests on the first benchmark path planning algorithm and the path planning algorithm to be compared in the different scenarios, respectively, to obtain the corresponding simulation results, and to evaluate the path planning algorithm to be compared based on the simulation results corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared.

[0037] A third aspect of this application provides an autonomous driving simulation test device, the device including a processor and a memory;

[0038] The memory is used to store program code and transmit the program code to the processor;

[0039] The processor is used to execute any one of the autonomous driving simulation test methods described in the first aspect according to the instructions in the program code.

[0040] A fourth aspect of this application provides a computer-readable storage medium for storing program code that, when executed by a processor, implements the autonomous driving simulation test method described in any of the first aspects.

[0041] As can be seen from the above technical solutions, this application has the following advantages:

[0042] This application provides an autonomous driving simulation testing method. The method involves segmenting collected complete road segment data into data fragments. For each frame of data in a data fragment, a first baseline path planning algorithm and a path planning algorithm to be compared are run to obtain the corresponding path planning trajectory for a future preset time period. By comparing the path planning trajectory corresponding to the path planning algorithm to be compared with that corresponding to the first baseline path planning algorithm, data frames with differences are identified, indicating differences caused by the path planning algorithm itself. Then, a difference scenario is generated based on the difference frames in the target data fragment. Simulation tests are performed using both the first baseline path planning algorithm and the path planning algorithm to be compared in the difference scenario to more specifically test the changes in autonomous driving capabilities caused by the iteration of the path planning algorithm to be compared. The simulation results of the first baseline path planning algorithm and the path planning algorithm to be compared in the difference scenario can then be used to evaluate the path planning algorithm to be compared. This allows for the evaluation of changes in the path planning algorithm to be compared through simulation testing, improving upon the technical problem that existing technologies cannot test changes in autonomous driving capabilities caused by path planning algorithm iteration in simulation testing. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 A flowchart illustrating an autonomous driving simulation testing method provided in this application embodiment;

[0045] Figure 2 This is a schematic diagram of an autonomous driving simulation testing device provided in an embodiment of this application. Detailed Implementation

[0046] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0047] For easier understanding, please refer to Figure 1 This application provides an autonomous driving simulation testing method, including:

[0048] Step 101: Divide the collected complete road segment data into several data fragments.

[0049] The collected complete road segment data can be the complete road segment data collected by autonomous vehicles during actual road testing. To facilitate simulation testing, the collected complete road segment data can be segmented. The complete road segment data can be divided into several data segments at preset time intervals (such as two minutes), and the resulting smaller data segments are easier to store and manage.

[0050] Step 102: Input each frame of data in the data segment into the first benchmark path planning algorithm and the path planning algorithm to be compared, respectively, to perform path planning and obtain the path planning trajectory for the corresponding future preset time period. Based on the path planning trajectory corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared, determine whether there are difference frames in the data segment, and obtain the target data segment with difference frames.

[0051] The path planning algorithms to be compared are selected as the path planning algorithms to be compared. The path planning algorithms to be compared can be evaluated based on the first benchmark path planning algorithm. The number of path planning algorithms to be compared can be one or more.

[0052] Each frame of data in the data segment is input into the first benchmark path planning algorithm and the path planning algorithm to be compared, respectively, for path planning, to obtain the corresponding path planning trajectory for the future preset time period. It should be noted that no simulation test was performed at this stage; only each frame of data in the data segment was input into the first benchmark path planning algorithm and the path planning algorithm to be compared, respectively, for path planning, to obtain the corresponding path planning trajectory for the future preset time period output by the first benchmark path planning algorithm and the path planning algorithm to be compared. Each frame of data in this data segment corresponds to a path planning trajectory for the future preset time period output by the first benchmark path planning algorithm and the path planning algorithm to be compared.

[0053] After obtaining the path planning trajectories from each frame of data in a data segment by running the first baseline path planning algorithm and the path planning algorithm to be compared, it is possible to determine whether there are differing frames in the data segment by comparing the path planning trajectories corresponding to the first baseline path planning algorithm and the path planning algorithm to be compared, thus obtaining the target data segment containing differing frames. The difference value between the path planning trajectories output by the first baseline path planning algorithm and the path planning algorithm to be compared for the same frame of data can be calculated; when the difference value is greater than a preset distance threshold, the data frame corresponding to that difference value is determined to be a differing frame; the data segment containing the differing frame is then used as the target data segment. Frame-by-frame scanning and comparison of the path planning trajectories corresponding to the first baseline path planning algorithm and the path planning algorithm to be compared can be used to determine whether there are differing frames in the data segment. The distance between the path planning trajectories output by the first benchmark path planning algorithm and the path planning algorithm to be compared for the same frame of data can be calculated to obtain the difference value. Since the first benchmark path planning algorithm and the path planning algorithm to be compared output the path planning trajectory for a future preset time period for each frame of data, the distance between the path planning trajectory for the future preset time period output by the first benchmark path planning algorithm and the path planning algorithm to be compared for each frame of data for each time period can be calculated at each time point. Alternatively, some time points can be selected to calculate the distance, such as calculating the distance of the path planning trajectory for the future preset time period output by the first benchmark path planning algorithm and the path planning algorithm to be compared for each frame of data at the 1st second, the 3rd second, the 5th second, the 8th second, etc., to obtain the difference value of the two path planning trajectories at several time points. Then, the difference value at each time point is compared with the corresponding preset distance threshold. When the difference value at a certain time point is greater than the corresponding preset distance threshold, the data frame corresponding to the difference value is determined to be the difference frame, that is, the data frame whose difference is caused by the path planning algorithm to be compared itself.

[0054] For example, assuming that the first baseline path planning algorithm and the path planning algorithm to be compared output the path planning trajectory for the next 8 seconds for each frame of data, when calculating the difference value, the straight-line distance between the path planning trajectories output by the first baseline path planning algorithm and the path planning algorithm to be compared for the same frame of data at the 1st, 3rd, 5th, and 8th seconds can be calculated to obtain the difference value dist between the path planning trajectories output by the first baseline path planning algorithm and the path planning algorithm to be compared for the same frame of data at the 1st second. 1s The difference value at the 3rd second, dist 3s The difference value at the 5th second, dist 5s The difference value between the second and the 8th second is dist 8s If a certain frame of data satisfies dist 1s >0.5, or dist 3s >0.66, or dist 5s >0.8, or dist 8s If the value is greater than 1.25, then the data frame is determined to be a difference frame, meaning that the first baseline path planning algorithm and the path planning algorithm to be compared have differences in this data frame. Thus, the data segment containing the difference frame can be obtained, which is the target data segment.

[0055] Step 103: Group the difference frames in the target data segment to generate several difference scenes.

[0056] After obtaining the target data segment containing differing frames, the differing frames in the target data segment can be sorted in ascending order by time. The differing frames in the target data segment with a time difference less than a preset time threshold (such as 1 second) are merged to obtain the set of differing frames in the target data segment. The set of differing frames in the target data segment is then segmented to obtain several differing scenes. The difference set can be segmented in the form of a sliding window to generate several differing scenes. For example, the merged set of differing frames is scanned from left to right and accumulated sequentially. If the accumulated number of differing frames exceeds 6 frames, these accumulated frames are extracted as a differing scene, thus obtaining a scene where the differing points are clustered.

[0057] Step 104: In the difference scenario, the first benchmark path planning algorithm and the path planning algorithm to be compared are used for simulation test respectively to obtain the corresponding simulation results, and the path planning algorithm to be compared is evaluated based on the simulation results corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared.

[0058] In the differential scenario, the first baseline path planning algorithm is used for simulation testing. The process of obtaining the corresponding simulation results is as follows: the first frame of data in the differential scenario is used as the current frame and input into the first baseline path planning algorithm for path planning. The simulation vehicle is controlled to drive according to the path planning trajectory for the future preset time period obtained from the current frame. The next frame of data in the differential scenario is used as the current frame and input into the first baseline path planning algorithm for path planning. The simulation vehicle is controlled to drive according to the path planning trajectory for the future preset time period obtained from the current frame. This process continues until all data frames in the differential scenario are tested, and the simulation results of the first baseline path planning algorithm in the differential scenario are obtained.

[0059] In the difference scenario, the simulation test is carried out using the path planning algorithm to be compared, and the corresponding simulation results are obtained as follows: the first frame of data in the difference scenario is used as the current frame and input into the path planning algorithm to be compared for path planning, and the simulation vehicle is controlled to drive according to the path planning trajectory for the future preset time period obtained from the current frame; the next frame of data in the difference scenario is used as the current frame and input into the path planning algorithm to be compared for path planning, and the simulation vehicle is controlled to drive according to the path planning trajectory for the future preset time period obtained from the current frame, until all data frames in the difference scenario are tested, and the simulation results of the path planning algorithm to be compared in the difference scenario are obtained.

[0060] After obtaining the simulation results of the first benchmark path planning algorithm and the path planning algorithm to be compared in different scenarios, the path planning algorithm to be compared is evaluated based on the simulation results corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared. Specifically, the simulation results of the first benchmark path planning algorithm in each different scenario can be measured according to the selected scenario metrics to obtain a first metric value for each different scenario. The scenario metrics include scenario safety and / or scenario comfort, etc.; the simulation results of the path planning algorithm to be compared in each different scenario can be measured according to the selected scenario metrics to obtain a second metric value for each different scenario; the path planning algorithm to be compared is evaluated based on the first and second metric values ​​for each different scenario.

[0061] When measuring different scenarios based on simulation results, scenario safety can be measured based on the distance between the simulated vehicle and surrounding vehicles, whether a collision occurs, etc., in the simulation test results. Scenario comfort can be measured based on the speed changes, acceleration / deceleration, and whether there is sudden braking when the simulated vehicle changes lanes during the simulation test. If there are multiple selected scenario measurement indicators, after measuring each indicator for each different scenario based on the simulation results of the first baseline path planning algorithm and the path planning algorithm to be compared in each different scenario, the measurement results of each indicator for the same different scenario can be weighted and summed to obtain the first and second measurement values ​​for each different scenario.

[0062] The algorithm compares the first and second metrics for the same difference scenario. When the second metric for the same difference scenario is greater than the corresponding first metric, the scenario is considered an improved scenario. When the second metric for the same difference scenario is less than the corresponding first metric, the scenario is considered a worsening scenario, i.e., a worsening scenario caused by the iteration of the path planning algorithm itself. Improved scenarios indicate that the path planning algorithm performs well in this type of scenario, while worsening scenarios indicate that it performs poorly. The algorithm is evaluated based on the number of improved and worsening scenarios. If the number of improved scenarios is greater than the number of worsening scenarios, the algorithm performs better than the benchmark algorithm. If the number of improved scenarios is less than the number of worsening scenarios, the algorithm performs worse than the benchmark algorithm. By comparing the first and second metrics for the same difference scenario, improved and worsening scenarios are identified, thus testing the changes in autonomous driving capabilities caused by the iteration of the path planning algorithm itself. This allows for a more comprehensive test of the performance of the path planning algorithm being compared.

[0063] Furthermore, after evaluating the path planning algorithm to be compared based on the simulation results corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared, a second benchmark path planning algorithm can be used for simulation testing in different scenarios, and the evaluation results obtained by the path planning algorithm to be compared can be verified based on the simulation results obtained by the second benchmark path planning algorithm.

[0064] To more accurately evaluate the path planning algorithms under comparison, a second benchmark path planning algorithm can be used for simulation testing in different scenarios. The simulation results can be measured using selected scenario metrics to obtain a third metric for each scenario. Scenario metrics include scenario safety and / or scenario comfort. The third and second metric values ​​for the same scenario can be compared. When the second metric value for the same scenario is greater than the corresponding third metric value, the scenario is considered an improved scenario. When the second metric value is less than the corresponding third metric value, the scenario is considered a worsening scenario, i.e., a worsening scenario caused by the iteration of the path planning algorithm itself. The path planning algorithm under comparison is evaluated based on the number of improved and worsening scenarios. If the evaluation results obtained using the simulation results from the second benchmark path planning algorithm are basically consistent with the evaluation results obtained using the simulation results from the first benchmark path planning algorithm, then the evaluation results of the path planning algorithm under comparison are verified to be reliable.

[0065] In this embodiment, the collected complete road segment data is segmented into data fragments. For each frame of data in a data fragment, a first baseline path planning algorithm and a path planning algorithm to be compared are run to obtain the corresponding path planning trajectory for a future preset time period. By comparing the path planning trajectory corresponding to the path planning algorithm to be compared with the path planning trajectory corresponding to the first baseline path planning algorithm, data frames with differences are identified, i.e., the differences caused by the path planning algorithm to be compared itself are determined. Then, a difference scenario is generated based on the difference frames in the target data fragment. In the difference scenario, simulation tests are performed using the first baseline path planning algorithm and the path planning algorithm to be compared, respectively, to more specifically test the changes in autonomous driving capabilities caused by the iteration of the path planning algorithm to be compared. The path planning algorithm to be compared can then be evaluated based on the simulation results of the first baseline path planning algorithm and the path planning algorithm to be compared in the difference scenario. This allows for the evaluation of changes in the path planning algorithm to be compared through simulation testing, improving the technical problem that existing technologies cannot test changes in autonomous driving capabilities caused by path planning algorithm iteration in simulation testing.

[0066] The above is an embodiment of an autonomous driving simulation testing method provided by this application. The following is an embodiment of an autonomous driving simulation testing device provided by this application.

[0067] Please refer to Figure 2 The autonomous driving simulation testing device provided in this application embodiment is characterized by comprising:

[0068] The segmentation unit is used to segment the collected complete road segment data into several data fragments;

[0069] The path planning unit is used to input each frame of data in the data segment into the first benchmark path planning algorithm and the path planning algorithm to be compared, respectively to perform path planning, obtain the path planning trajectory for the corresponding future preset time period, and determine whether there are difference frames in the data segment based on the path planning trajectories corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared, and obtain the target data segment with difference frames.

[0070] The generation unit is used to group the difference frames in the target data segment and generate several difference scenarios;

[0071] The simulation unit is used to perform simulation tests on the first benchmark path planning algorithm and the path planning algorithm to be compared in different scenarios, obtain the corresponding simulation results, and evaluate the path planning algorithm to be compared based on the simulation results of the first benchmark path planning algorithm and the path planning algorithm to be compared.

[0072] As a further improvement, the path planning unit is specifically used for:

[0073] Each frame of data in the data segment is input into the first baseline path planning algorithm and the path planning algorithm to be compared, respectively, to perform path planning and obtain the path planning trajectory for the corresponding future preset time period.

[0074] Calculate the difference between the path planning trajectories output by the first baseline path planning algorithm and the path planning algorithm to be compared for the same frame of data;

[0075] When the difference value is greater than the preset distance threshold, the data frame corresponding to the difference value is determined to be a difference frame.

[0076] Use the data fragments of the frames with differences as the target data fragments.

[0077] As a further improvement, the generation unit is specifically used for:

[0078] Sort the difference frames in the target data segment in ascending order by time;

[0079] Merge the difference frames in the target data segment whose time difference is less than a preset time threshold to obtain the difference frame set of the target data segment;

[0080] The set of difference frames of the target data segment is divided into several difference scenes.

[0081] As a further improvement, the simulation unit is specifically used for:

[0082] The first frame data in the different scenarios is used as the current frame and input into the first benchmark path planning algorithm and the path planning algorithm to be compared respectively for path planning. The simulation vehicle is controlled to drive according to the path planning trajectory of the future preset time period obtained from the current frame.

[0083] The next frame data in the difference scenario is used as the current frame and input into the first benchmark path planning algorithm and the path planning algorithm to be compared respectively for path planning. The simulation vehicle is controlled to drive according to the path planning trajectory of the future preset time period obtained from the current frame until all data frames in the difference scenario are tested, and the simulation results of the first benchmark path planning algorithm and the path planning algorithm to be compared in the difference scenario are obtained.

[0084] Based on the selected scenario metrics, the simulation results of the first baseline path planning algorithm in each different scenario are measured to obtain the first metric value for each different scenario. The scenario metrics include scenario safety and / or scenario comfort.

[0085] Based on the selected scenario metrics, the simulation results of the path planning algorithm to be compared are measured in various different scenarios to obtain a second metric value for each different scenario.

[0086] The path planning algorithms to be compared are evaluated using the first and second metrics for each different scenario.

[0087] As a further improvement, it also includes:

[0088] The verification unit is used to perform simulation tests using the second benchmark path planning algorithm in different scenarios, and to verify the evaluation results obtained by the path planning algorithm to be compared based on the simulation results obtained by the second benchmark path planning algorithm.

[0089] In this embodiment, the collected complete road segment data is segmented into data fragments. For each frame of data in a data fragment, a first baseline path planning algorithm and a path planning algorithm to be compared are run to obtain the corresponding path planning trajectory for a future preset time period. By comparing the path planning trajectory corresponding to the path planning algorithm to be compared with the path planning trajectory corresponding to the first baseline path planning algorithm, data frames with differences are identified, i.e., the differences caused by the path planning algorithm to be compared itself are determined. Then, a difference scenario is generated based on the difference frames in the target data fragment. In the difference scenario, simulation tests are performed using the first baseline path planning algorithm and the path planning algorithm to be compared, respectively, to more specifically test the changes in autonomous driving capabilities caused by the iteration of the path planning algorithm to be compared. The path planning algorithm to be compared can then be evaluated based on the simulation results of the first baseline path planning algorithm and the path planning algorithm to be compared in the difference scenario. This allows for the evaluation of changes in the path planning algorithm to be compared through simulation testing, improving the technical problem that existing technologies cannot test changes in autonomous driving capabilities caused by path planning algorithm iteration in simulation testing.

[0090] This application also provides an autonomous driving simulation test device, which includes a processor and a memory;

[0091] The memory is used to store program code and transfer the program code to the processor;

[0092] The processor is used to execute the autonomous driving simulation test method in the foregoing method embodiments according to the instructions in the program code.

[0093] This application also provides a computer-readable storage medium for storing program code, which, when executed by a processor, implements the autonomous driving simulation test method in the aforementioned method embodiments.

[0094] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and unit can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0095] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0096] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0097] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0098] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0099] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0100] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for executing all or part of the steps of the methods described in the various embodiments of this application through a computer device (which may be a personal computer, server, or network device, etc.). The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0101] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An autonomous driving simulation testing method, characterized in that, include: The collected complete road segment data is divided into several data fragments; Each frame of data in the data segment is input into the first benchmark path planning algorithm and the path planning algorithm to be compared for path planning, respectively, to obtain the path planning trajectory for the corresponding future preset time period. Based on the path planning trajectory corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared, it is determined whether there are difference frames in the data segment, and the target data segment with difference frames is obtained. The difference frames in the target data segment are grouped to generate several difference scenarios, including: Sort the difference frames in the target data segment in ascending order by time; Merge the difference frames in the target data segment whose time difference is less than a preset time threshold to obtain the difference frame set of the target data segment; The set of difference frames of the target data segment is divided into several difference scenarios; In the aforementioned difference scenario, the first benchmark path planning algorithm and the path planning algorithm to be compared are used for simulation testing to obtain corresponding simulation results. The path planning algorithm to be compared is evaluated based on the simulation results of the first benchmark path planning algorithm and the path planning algorithm to be compared.

2. The autonomous driving simulation testing method according to claim 1, characterized in that, The step of determining whether there are difference frames in the data segment based on the path planning trajectories corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared, and obtaining the target data segment with difference frames, includes: Calculate the difference between the path planning trajectories output by the first baseline path planning algorithm and the path planning algorithm to be compared for the same frame of data; When the difference value is greater than a preset distance threshold, the data frame corresponding to the difference value is determined to be a difference frame; The data segment containing the differing frame is taken as the target data segment.

3. The autonomous driving simulation testing method according to claim 1, characterized in that, In the aforementioned difference scenario, simulation tests are performed using the first baseline path planning algorithm and the path planning algorithm to be compared, respectively, to obtain corresponding simulation results, including: The first frame data in the difference scenario is used as the current frame and input into the first benchmark path planning algorithm and the path planning algorithm to be compared respectively for path planning. The simulation vehicle is controlled to drive according to the path planning trajectory of the future preset time period obtained from the current frame. The next frame data in the difference scenario is used as the current frame and input into the first benchmark path planning algorithm and the path planning algorithm to be compared for path planning. The simulation vehicle is controlled to drive according to the path planning trajectory of the future preset time period obtained from the current frame until all data frames in the difference scenario are tested, and the simulation results of the first benchmark path planning algorithm and the path planning algorithm to be compared in the difference scenario are obtained.

4. The autonomous driving simulation testing method according to claim 1 or 3, characterized in that, The step of evaluating the path planning algorithm to be compared based on the simulation results corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared includes: Based on the selected scenario metric, the simulation results of the first baseline path planning algorithm in each of the different scenarios are measured to obtain a first metric value for each of the different scenarios. The scenario metric includes scenario safety and / or scenario comfort. Based on the selected scenario metric, the simulation results of the path planning algorithm to be compared in each of the different scenarios are measured to obtain a second metric value for each of the different scenarios; The path planning algorithms to be compared are evaluated using the first and second metrics for each of the different scenarios.

5. The autonomous driving simulation testing method according to claim 4, characterized in that, The evaluation of the path planning algorithm to be compared using the first and second metrics for each of the aforementioned difference scenarios includes: Compare the first metric and the second metric for the same difference scenario; When the second metric value of the same difference scenario is greater than the corresponding first metric value, the difference scenario is determined to be an improved scenario; When the second metric value of the same difference scenario is less than the corresponding first metric value, the difference scenario is determined to be a variation scenario; The path planning algorithm to be compared is evaluated based on the number of improved scenarios and the number of worse scenarios.

6. The autonomous driving simulation testing method according to claim 1, characterized in that, The method further includes: In the aforementioned difference scenario, a second benchmark path planning algorithm is used for simulation testing, and the evaluation results obtained by the path planning algorithm to be compared are verified based on the simulation results obtained by the second benchmark path planning algorithm.

7. The autonomous driving simulation testing method according to claim 1, characterized in that, The number of path planning algorithms to be compared is one or more.

8. An autonomous driving simulation testing device, characterized in that, include: The segmentation unit is used to segment the collected complete road segment data into several data fragments; The path planning unit is used to input each frame of data in the data segment into the first benchmark path planning algorithm and the path planning algorithm to be compared, respectively to perform path planning, obtain the path planning trajectory for the corresponding future preset time period, and determine whether there are difference frames in the data segment based on the path planning trajectories corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared, so as to obtain the target data segment with difference frames. The generation unit is used to group the difference frames in the target data segment and generate several difference scenarios, including: Sort the difference frames in the target data segment in ascending order by time; Merge the difference frames in the target data segment whose time difference is less than a preset time threshold to obtain the difference frame set of the target data segment; The set of difference frames of the target data segment is divided into several difference scenarios; The simulation unit is used to perform simulation tests on the first benchmark path planning algorithm and the path planning algorithm to be compared in the different scenarios, respectively, to obtain the corresponding simulation results, and to evaluate the path planning algorithm to be compared based on the simulation results corresponding to the first benchmark path planning algorithm and the path planning algorithm to be compared.

9. An autonomous driving simulation testing device, characterized in that, The device includes a processor and a memory; The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the autonomous driving simulation test method according to any one of claims 1-7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code, which, when executed by a processor, implements the autonomous driving simulation test method according to any one of claims 1-7.

Images