Simulation test method and device of unmanned vehicle and storage medium

Abstract

The invention provides an unmanned vehicle simulation test method and device and a storage medium, and the method comprises the steps: carrying out the synchronous playback of operation records of log files of a plurality of unmanned vehicles, and extracting first scene information; in the playback process, first driving information is determined and shared in real time for each unmanned vehicle, and second driving information of surrounding unmanned vehicles is determined according to the shared first shared information under the condition that surrounding unmanned vehicles exist in a preset range; determining input information of a first algorithm according to the first scene information, the first driving information and the second driving information, and inputting the input information to the first algorithm; and adjusting the input information of the subsequent first algorithm according to the output of the first algorithm until playback is completed, and obtaining a test result.

Description

Technical Field

[0001] The present disclosure relates to the field of driverless technologies, and in particular, to a simulation test method, device, and storage medium for driverless vehicles. Background Art

[0002] In the scenario of driverless operation in mines, the operation of driverless vehicles involves continuous loading, unloading, crushing, etc. links, and they do not operate independently, but multiple driverless vehicles need to cooperate and work together. For example, driverless vehicles need to interact with surrounding driverless vehicles according to the actual situation, and make decisions on key matters such as the right of way and driving route to ensure the efficient and safe operation of mine operations. However, traditional simulation test methods only rely on the log files of single vehicles to restore the driving process and parking reasons of driverless vehicles. Summary of the Invention

[0003] Embodiments of the present disclosure provide a simulation test method, device, and storage medium for driverless vehicles to solve the problem that existing effective simulation tests cannot be performed for the scenario of cooperative operation of driverless vehicles.

[0004] Based on the above problems, in a first aspect, a simulation test method for driverless vehicles provided by embodiments of the present disclosure includes:

[0005] Obtain log files generated during the operation of multiple driverless vehicles participating in the simulation test;

[0006] Synchronously playback the operation records of the log files respectively corresponding to the multiple driverless vehicles, and extract first scenario information to simulate the historical operation scenarios of the multiple driverless vehicles;

[0007] During the playback process, for each driverless vehicle, according to the operation records of the log file, determine and share the first driving information of the driverless vehicle in real time, and when there are surrounding driverless vehicles within the preset range of the driverless vehicle, determine the second driving information of the surrounding driverless vehicles according to the first shared information shared by the surrounding driverless vehicles in real time;

[0008] According to the first scenario information, the first driving information, and the second driving information, determine the input information of the first algorithm corresponding to the driverless vehicle, and input the input information into the first algorithm;

[0009] According to the output of the first algorithm, adjust the input information of the subsequent first algorithm in real time until the playback is completed to obtain the test result of the first algorithm.

[0010] In combination with the first aspect, in a possible implementation manner, the synchronously playing back the operation records of the log files respectively corresponding to the multiple driverless vehicles, extracting first scenario information to simulate the historical operation scenarios of the multiple driverless vehicles includes:

[0011] Determine the playback frame rate according to the running record frame rates corresponding to each module included in the log file.

[0012] Align the running records of each module included in the log files corresponding to the multiple autonomous vehicles according to a preset timestamp, and control synchronous playback at the playback frame rate to extract first scene information.

[0013] Simulate the historical running scenarios of the multiple autonomous vehicles according to the first scene information.

[0014] Combined with the first aspect, in a possible implementation manner, when there are surrounding autonomous vehicles within the preset range of the autonomous vehicle, it further includes:

[0015] Determine first detection information of the surrounding autonomous vehicles according to the running records of the log file; and

[0016] Determine whether the surrounding autonomous vehicles belong to the multiple autonomous vehicles according to the identification information corresponding to the surrounding autonomous vehicles in the first detection information.

[0017] The determining the second driving information of the surrounding autonomous vehicles according to the first shared information shared by the surrounding autonomous vehicles in real time includes:

[0018] If it belongs, determine the first shared information shared by the surrounding autonomous vehicle in real time as the second driving information of the surrounding autonomous vehicle; or,

[0019] If it belongs, compare the first detection information with the first shared information shared by the surrounding autonomous vehicle in real time; in the case where the comparison result shows that the first detection information is inconsistent with the first shared information, determine the first shared information as the second driving information of the surrounding autonomous vehicle.

[0020] Combined with the first aspect, in a possible implementation manner, it further includes:

[0021] If it belongs and the comparison result is consistent, determine the first detection information as the second driving information of the surrounding autonomous vehicle; and / or,

[0022] If it does not belong, determine the first detection information as the second driving information of the surrounding autonomous vehicle.

[0023] Combined with the first aspect, in a possible implementation manner, the real-time adjustment of the input information of the subsequent first algorithm according to the output of the first algorithm includes:

[0024] Obtain new first driving information of the autonomous vehicle included in the output of the current first algorithm.

[0025] Share by using the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and

[0026] Determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information.

[0027] Combined with the first aspect, in a possible implementation manner, after obtaining the new first driving information, it further includes:

[0028] Compare the new first driving information with the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle;

[0029] Share by using the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information, including:

[0030] In the case where the comparison results are inconsistent, share by using the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information;

[0031] The method further includes:

[0032] In the case where the comparison results are consistent, share the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the first driving information, and the second driving information.

[0033] Combined with the first aspect, in a possible implementation manner, the following method is adopted to obtain the log files generated during the operation of multiple driverless vehicles participating in the simulation test:

[0034] Obtain the log files generated during the operation of multiple driverless vehicles participating in the simulation test based on the method of real vehicle latch recording; and / or,

[0035] Obtain the log files generated during the operation of multiple driverless vehicles participating in the simulation test based on the method of simulation recording.

[0036] Combined with the first aspect, in a possible implementation manner, the first algorithm includes a second algorithm corresponding to the operation record of the log file or a third algorithm updated based on the second algorithm.

[0037] Second aspect, there is provided a simulation test device for driverless vehicles, including:

[0038] A log acquisition module, configured to acquire log files generated during the operation of multiple driverless vehicles participating in the simulation test;

[0039] An input information determination module, configured to synchronously replay the operation records of the log files respectively corresponding to the multiple driverless vehicles, extract first scenario information, so as to simulate the historical operation scenarios of the multiple driverless vehicles; and

[0040] During the replay process, for each driverless vehicle, according to the operation records of the log file, the first driving information of the driverless vehicle is determined and shared in real time, and when there are surrounding driverless vehicles within the preset range of the driverless vehicle, according to the first shared information shared in real time by the surrounding driverless vehicles, the second driving information of the surrounding driverless vehicles is determined;

[0041] According to the first scenario information, the first driving information and the second driving information, the input information corresponding to the first algorithm of the driverless vehicle is determined, and the input information is input into the first algorithm;

[0042] An output module, configured to adjust the input information of the subsequent first algorithm in real time according to the output of the first algorithm until the replay is completed, and obtain the test result of the first algorithm.

[0043] Third aspect, there is provided a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the steps of the simulation test method for driverless vehicles as described in the first aspect or any possible implementation manner in combination with the first aspect.

[0044] The beneficial effects of the embodiments of the present disclosure include: A simulation test method, device, and storage medium for an autonomous vehicle provided by the present disclosure, including: obtaining log files generated during the operation of multiple autonomous vehicles participating in the simulation test; synchronously playing back the operation records of the log files respectively corresponding to the multiple autonomous vehicles to extract first scenario information for simulating the historical operation scenarios of the multiple autonomous vehicles; during the playback process, for each autonomous vehicle, according to the operation records of the log file, determining and sharing in real time the first driving information of the autonomous vehicle, and in the case where there are surrounding autonomous vehicles within the preset range of the autonomous vehicle, determining the second driving information of the surrounding autonomous vehicles according to the first shared information shared in real time by the surrounding autonomous vehicles; according to the first scenario information, the first driving information, and the second driving information, determining the input information of the first algorithm corresponding to the autonomous vehicle and inputting the input information into the first algorithm; adjusting the input information of the subsequent first algorithm in real time according to the output of the first algorithm until the playback is completed to obtain the test result of the first algorithm. Compared with the prior art, the simulation test method for an autonomous vehicle provided by the embodiments of the present disclosure parses the log files and synchronously plays back the operation records of the log files respectively corresponding to each autonomous vehicle in the collaborative operation scenario, thereby simulating the historical operation scenarios of each autonomous vehicle, and sharing driving information in real time among each autonomous vehicle during the playback process. Therefore, the simulation test results fully consider the problems of autonomous vehicle collaborative operation, making the simulation test results more adaptable to the real driverless scenario, especially large-scale driverless clusters. Furthermore, it also enables the test simulation to better support the optimization of the driverless algorithm for large-scale driverless clusters. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flowchart of the simulation test method for an autonomous vehicle provided by the embodiments of the present disclosure;

[0046] Figure 2 It is one of the schematic diagrams of the simulation test method for an autonomous vehicle provided by the embodiments of the present disclosure;

[0047] Figure 3 It is another schematic diagram of the simulation test method for an autonomous vehicle provided by the embodiments of the present disclosure;

[0048] Figure 4 It is one of the schematic diagrams of the simulation test results for an autonomous vehicle provided by the embodiments of the present disclosure;

[0049] Figure 5 It is another schematic diagram of the simulation test results for an autonomous vehicle provided by the embodiments of the present disclosure;

[0050] Figure 6 It is the structural diagram of the simulation test device for an autonomous vehicle provided by the embodiments of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] Embodiments of the present disclosure provide a simulation test method, apparatus, and storage medium for an autonomous vehicle. The preferred embodiments of the present disclosure will be described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present disclosure and are not used to limit the present disclosure. And without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0052] Embodiments of the present disclosure provide a simulation test method for an autonomous vehicle, as Figure 1 shown, including the following steps:

[0053] S101. Obtain log files generated during the operation of multiple autonomous vehicles participating in the simulation test;

[0054] S102. Synchronously replay the operation records of the log files respectively corresponding to the multiple autonomous vehicles, and extract first scenario information to simulate the historical operation scenarios of the multiple autonomous vehicles;

[0055] S103. During the replay process, for each autonomous vehicle, according to the operation records of the log file, determine and share the first driving information of the autonomous vehicle in real time, and when there are surrounding autonomous vehicles within the preset range of the autonomous vehicle, determine the second driving information of the surrounding autonomous vehicles according to the first shared information shared by the surrounding autonomous vehicles in real time;

[0056] S104. Determine the input information of the first algorithm corresponding to the autonomous vehicle according to the first scenario information, the first driving information, and the second driving information, and input the input information into the first algorithm;

[0057] S105. Adjust the input information of the subsequent first algorithm in real time according to the output of the first algorithm until the replay is completed to obtain the test result of the first algorithm.

[0058] In the embodiments of the present disclosure, in the field of autonomous driving technology, a log simulation (logsim) tool can be used for the development and testing of autonomous vehicles. By playing back the log files generated during the operation of an autonomous vehicle, problems such as the decision-making or performance of the autonomous vehicle can be reproduced and regression tested. Using the logsim tool can quickly and efficiently verify the reliability of algorithms without relying on costly physical tests. The log files can include sensor data recorded during the actual operation of the autonomous vehicle, such as data collected in real time by cameras, lidar, radar, GPS systems, etc., and can also include the responses of algorithm modules to the processing of these sensor data, such as the speed, position, driving direction, driving trajectory, information about obstacles ahead, and operation instructions of the autonomous vehicle. These operation records are organized into detailed log files, providing the basic materials for subsequent simulation using the logsim tool. When using logsim for simulation testing, flow control technology can control the playback process of the operation records in the log file to ensure that data is input into the simulation module and algorithm module at an appropriate speed and order, aiming to ensure the authenticity, real-time nature, and stability of the simulation process, avoid data overload or loss, and thus accurately conduct simulation testing on the autonomous vehicle.

[0059] Based on the combination of logsim and flow control technology, the driving process and parking reasons of the autonomous vehicle in the log file can be accurately restored, which is of great benefit to the analysis of single-vehicle abnormal problems. However, when dealing with problems related to the collaborative operation of autonomous vehicles, such as the collaborative operation of a large-scale autonomous driving cluster, simply relying on the logsim playback reproduction of single-vehicle log files is not sufficient to complete the analysis and solution of the problems. For example, during the process of dealing with problems related to the collaborative operation of autonomous vehicles, the playback processes of log files among multiple autonomous vehicles are independent of each other, with inconsistent timings, lack of real-time information of surrounding autonomous vehicles, and inability to update the decision-making strategy of the vehicle itself using the information of surrounding autonomous vehicles. These defects all have a huge impact on the analysis and solution of problems related to the collaborative operation of autonomous vehicles.

[0060] In the embodiments of the present disclosure, log files generated during the operation of multiple autonomous vehicles participating in simulation tests are obtained. These autonomous vehicles assist each other during operation and participate in driving decisions. For example, these autonomous vehicles interact to decide on the right of way and driving routes through vehicle-to-everything (V2X) communication technology. Each autonomous vehicle generates a log file during operation. The operation records of each log file are synchronously replayed. Flow control technology can be used to register multiple autonomous vehicles with a simulation flow control service, enabling synchronous replay among multiple autonomous vehicles. For example, if multiple autonomous vehicles include autonomous vehicles a, b, and c, they are synchronously replayed according to the timestamp records in the log files. When abnormal situations such as lag and delay occur during the replay of autonomous vehicle a, autonomous vehicles b and c should stop and wait. After the replay of autonomous vehicle a resumes normal, autonomous vehicles b and c continue to replay.

[0061] Furthermore, registering the simulation flow control service for each module corresponding to the log file records enables synchronous replay among the various modules within the autonomous vehicle. For example, when synchronously replaying the data collected by the camera module and the lidar module in the log file records, when abnormal situations such as lag and delay occur during the replay of the camera module, the replay of the lidar module should stop and wait. After the replay of the camera module resumes normal, the lidar module continues to replay. This can ensure synchronous replay of the operation records of the log files respectively corresponding to multiple autonomous vehicles.

[0062] In the embodiments of the present disclosure, the scene information in the log file can be extracted as the first scene information. Exemplarily, the first scene information may include infrastructure information of the driving road where the autonomous vehicle is located, such as lane line information, traffic sign / traffic signal information, boundary information such as roadside isolation belts, and obstacle information. The first scene information may also include natural environment information, such as rain / snow weather information and visibility information corresponding to the time of the log record. By using the extracted first scene information, the scene where the autonomous vehicle is located can be built, thereby simulating the historical operation scenes of multiple autonomous vehicles. For example, the extracted first scene information is input into the simulation module of logsim for scene building.

[0063] During the playback process, for each driverless vehicle, the running record content in the log file is extracted to obtain the first driving information of the driverless vehicle, which is then shared to the shared data storage. The first driving information may include real-time position information, real-time vehicle speed, blind spots, detected obstacle information, driving trajectory planning, and other information when the driverless vehicle is running. When there are surrounding driverless vehicles within a preset range of the position where the driverless vehicle is located or within a preset collaborative operation area range, the first shared information shared by the surrounding driverless vehicles is obtained from the shared data storage. From the perspective of the surrounding driverless vehicles, the above-mentioned first driving information shared to the shared data storage can be used as the first shared information. The first shared information may include: real-time position information, real-time vehicle speed, blind spots, detected obstacle information, driving trajectory planning, and other information when the surrounding driverless vehicles are running. By parsing the first shared information shared by the surrounding driverless vehicles in real time, it can be determined whether the driverless vehicle corresponding to the first shared information belongs to the multiple driverless vehicles participating in the simulation test, and the running state of the surrounding driverless vehicles can be obtained more accurately, avoiding the lack of real-time running state information of the surrounding driverless vehicles. In the case of belonging, the first shared information is determined as the second driving information of the surrounding driverless vehicles. Using the second driving information of the surrounding driverless vehicles to update the decision-making strategy of the own vehicle can obtain more accurate simulation test results.

[0064] Further, according to the first scenario information, the first driving information of the driverless vehicle, and the second driving information of the surrounding driverless vehicles, they are used as the input information of the first algorithm corresponding to the driverless vehicle. The first algorithm is reproduced for problems or regression tested. The call frame rate of the first algorithm is synchronized with the playback of the running records in the log file, and the input information is input into the first algorithm. The first algorithm may include various algorithms corresponding to each module in the log file. When the first algorithm optimizes the decision-making strategy of the driverless vehicle, such as the optimization of the driverless algorithm for a large-scale driverless cluster, the output of the first algorithm may be different from some records in the log file. For example, the output of the first algorithm includes the re-planning of the driving trajectory of the driverless vehicle. Therefore, during the playback process, the input information of the subsequent first algorithm is adjusted in real time according to the output of the first algorithm until the playback is completed, so as to obtain the test result of the first algorithm. Exemplarily, as Figure 2 shown, the log files of the multiple driverless vehicles obtained are rosbagfile1, rosbagfile2, and rosebagfile3 respectively. The flow control technology is used for synchronous playback. During the playback process, the logsim tool is used to extract the first scenario information in the log file for simulating the historical running scenarios of the multiple driverless vehicles. The first driving information is shared to the shared data storage, and the first shared information of the surrounding driverless vehicles is obtained from the shared data storage. After parsing the first shared information, the second driving information of the surrounding driverless vehicles is determined, and the first scenario information, the first driving information, and the second driving information are input into the first algorithm for testing.

[0065] In the embodiment of the present application, by parsing the log files corresponding to each autonomous vehicle in the collaborative operation scenario and synchronously playing back its operation records, the historical operation scenarios of each autonomous vehicle are restored. Compared with the prior art, during the playback process, the operation record playbacks of each log file are kept synchronized, and the autonomous vehicles share driving information in real time, so that the simulation test results fully consider the problems of autonomous vehicle collaborative operation, making the simulation test results more adaptable to the real driverless scenario, especially for large-scale driverless clusters. Furthermore, the test simulation can better support the optimization of the driverless algorithms for large-scale driverless clusters.

[0066] In another embodiment of the present disclosure, in the above step S102, synchronously playing back the operation records of the log files respectively corresponding to multiple autonomous vehicles to extract the first scenario information for simulating the historical operation scenarios of the multiple autonomous vehicles includes the following steps:

[0067] Step 1: Determine the playback frame rate according to the operation record frame rates respectively corresponding to each module included in the log file;

[0068] Step 2: Align the operation records of each module included in the log files respectively corresponding to multiple autonomous vehicles according to a preset time stamp, and control the playback frame rate for synchronous playback to extract the first scenario information;

[0069] Step 3: Simulate the historical operation scenarios of multiple autonomous vehicles according to the first scenario information.

[0070] In the embodiment of the present disclosure, by applying the flow control technology, the playback frame rate of the operation records respectively corresponding to each module in the log file is made consistent with the frame rate of each module during the actual operation of the autonomous vehicle, and the playback frame rate is controlled for synchronous playback among multiple autonomous vehicles, so as to ensure that the real scenario can be accurately reproduced through the log file. For example, the vehicle behaviors, environmental changes, event triggers, etc. in multi-vehicle collaborative operation are reproduced. For the above step 1, the log file usually includes time stamp information and the message types generated corresponding to the operation of each module. Extract the time stamp information in the log file and store it according to the message type. The frame rate of the operation records respectively corresponding to each module can be obtained according to the time interval of the operation records of each module in the log file, so as to obtain the playback frame rate and ensure that the playback frame rate is consistent with the frame rate of each module during the actual operation of the autonomous vehicle. For the above step 2, align the operation records corresponding to each module according to a preset time stamp. To ensure that the operation records of each module included in the log files respectively corresponding to multiple autonomous vehicles are synchronously played back in the simulation, the playback frame rate is controlled.

[0071] Exemplarily, during the actual driving process of the driverless vehicle, the camera, lidar, and radar collect data simultaneously. The camera collects images at a frequency of 25 Hz, the lidar collects point clouds at a frequency of 10 Hz, and the radar collects target information at a frequency of 20 Hz. The playback frame rate is determined according to the running record frame rates corresponding to each module, resulting in a camera image being played back every 40 ms, a lidar point cloud being played back every 100 ms, and radar data being played back every 50 ms. The running records of the camera, lidar, and radar are aligned according to the time stamps. For example, at the time stamp t = 1 s, the playback is synchronized according to the running order and time interval between the camera, lidar, and radar in the log record, avoiding any situation where the operation of any module is delayed. For example, when playing back the image of the camera at t = 5 s, the situation of playing back the point cloud of the lidar at t = 2 s is avoided.

[0072] Exemplarily, during the actual driving process of the driverless vehicle, in the running record content corresponding to the vehicle speed control module and the driving trajectory planning module, the frame rate of the running record of the vehicle speed control module is 25 Hz, and the frame rate of the driving trajectory planning module is 20 Hz. The playback frame rate is determined according to the running record frame rates corresponding to each module, resulting in a running record of the vehicle speed control module being played back every 40 ms and a running record of the driving trajectory planning module being played back every 50 ms. When any module has abnormal situations such as stuttering and delay during the playback process, other modules should stop and wait. After the playback of this module returns to normal, other modules continue to play back, thereby controlling the playback frame rate and achieving synchronous playback.

[0073] For problems related to the collaborative operation of driverless vehicles, the playback frame rates are controlled for synchronous playback among multiple driverless vehicles. The running records of each module included in the log files corresponding to multiple driverless vehicles are aligned according to a preset time stamp, and the playback frame rate is controlled for synchronous playback, thereby realizing the simulation of the historical operation scenarios of multiple driverless vehicles. Exemplarily, in the collaborative operation scenario of mine driverless vehicles, when driverless vehicle a completes the loading task and drives away from the loading position, at this time, driverless vehicles b and c enter the loading area and go to the loading position and the position to be loaded respectively. During the playback process, the running records of each module included in the log files corresponding to each driverless vehicle are aligned according to a preset time stamp, such as t = 1 s, and the playback frame rate is controlled for playback, avoiding the situation where when playing back that driverless vehicle a has driven away from the loading position at t = 5 s, the situation where driverless vehicle b is going to the loading position at t = 2 s is played back.

[0074] During the playback process, the first scenario information in the log file is extracted. The first scenario information may include infrastructure information of the road where the driverless vehicle is traveling, such as lane line information, traffic sign / signal information, boundary information such as roadside isolation belts, and obstacle information. The first scenario information may also include natural environment information, such as rain / snow weather information and visibility information at the time corresponding to the log record. For step 3 above, the simulation module of logsim can simulate the data of the operation records of each module in the log file and provide a clear and intuitive effect display. By using the simulation module of logsim to parse the extracted first scenario information, the scenario where the driverless vehicle is located can be built, so as to simulate the historical operation scenarios of multiple driverless vehicles. By controlling the operation records of the log files corresponding to multiple driverless vehicles to be played back in the correct time sequence and rate, the historical operation scenarios of multiple driverless vehicles can be simulated, and it is ensured that the input data to the first algorithm is consistent with the data in the actual scenario.

[0075] In another embodiment of the present disclosure, when there are surrounding driverless vehicles within the preset range of the driverless vehicle, it further includes:

[0076] Step 1: Determine the first detection information of the surrounding driverless vehicles according to the operation records of the log file;

[0077] Step 2: And determine whether the surrounding driverless vehicles belong to multiple driverless vehicles according to the identification information of the surrounding driverless vehicles corresponding in the first detection information;

[0078] In the above step S102, determining the second driving information of the surrounding driverless vehicles according to the first shared information shared by the surrounding driverless vehicles in real time includes:

[0079] Step 3: If it belongs, determine the first shared information shared by the surrounding driverless vehicles in real time as the second driving information of the surrounding driverless vehicles; or,

[0080] Step 4: If it belongs, compare the first detection information with the first shared information shared by the surrounding driverless vehicles in real time; in the case where the comparison result shows that the first detection information is inconsistent with the first shared information, determine the first shared information as the second driving information of the surrounding driverless vehicles.

[0081] In the embodiments of the present disclosure, the first shared information shared in real time by the surrounding driverless vehicles is obtained from the shared data storage according to the identification information of the surrounding driverless vehicles in the log file. For the above step 1, during the actual operation of the driverless vehicle, when there is an information interaction behavior between the driverless vehicle and the surrounding driverless vehicles or the driverless vehicle detects the existence of surrounding driverless vehicles, corresponding operation records will be generated in the log file. During the playback of the operation records in the log file, the first detection information of the surrounding driverless vehicles is extracted. These first detection information may include: the identification information of the surrounding driverless vehicles, the real-time position information, real-time vehicle speed, blind spots, detected obstacle information, and driving trajectory planning information of the surrounding driverless vehicles during operation. For the above step 2, it can be determined whether the surrounding driverless vehicles belong to multiple driverless vehicles by checking whether the identification information of the corresponding surrounding driverless vehicles in the first detection information is consistent with the identification information of multiple driverless vehicles participating in the simulation test. For the above step 3, when the surrounding driverless vehicles belong to multiple driverless vehicles, the first shared information shared by the surrounding driverless vehicles corresponding to the identification information can be obtained from the shared data storage. The first shared information is determined as the second driving information of the surrounding driverless vehicles. For the above step 4, when the surrounding driverless vehicles belong to multiple driverless vehicles, the first shared information shared by the surrounding driverless vehicles corresponding to the identification information is compared with the above first detection information. It can be compared whether the real-time position information, real-time vehicle speed, blind spots, detected obstacle information, and driving trajectory planning information of the surrounding driverless vehicles during operation are consistent. When the comparison result shows that the first detection information is inconsistent with the first shared information, the first shared information is determined as the second driving information of the surrounding driverless vehicles. By updating the decision-making strategy of the vehicle itself with the second driving information of the surrounding driverless vehicles determined by using the first shared information shared by the surrounding driverless vehicles, more accurate simulation test results can be obtained.

[0082] In another embodiment of the present disclosure, it further includes:

[0083] Step 1, if it belongs and the comparison result is consistent, determine the first detection information as the second driving information of the surrounding driverless vehicles; and / or,

[0084] Step 2, if it does not belong, determine the first detection information as the second driving information of the surrounding driverless vehicles.

[0085] In the embodiments of the present disclosure, the surrounding unmanned vehicles recorded in the log file may not belong to the multiple unmanned vehicles participating in the simulation test. For the above step 1, in the case where the surrounding unmanned vehicle belongs to the multiple unmanned vehicles, the first shared information shared by the surrounding unmanned vehicle corresponding to the identification information is compared with the first detection information, and it can be compared whether information such as the real-time position information, real-time vehicle speed, blind area, detected obstacle information, and driving trajectory planning is consistent when the surrounding unmanned vehicle is running. When the comparison result shows that the first detection information is consistent with the first shared information, the first detection information is determined as the second driving information of the surrounding unmanned vehicle. This can reduce the replacement operation of the first detection information of the surrounding unmanned vehicle extracted from the log file and improve the simulation efficiency. For the above step 2, the surrounding unmanned vehicle recorded in the log file does not belong to the multiple unmanned vehicles participating in the simulation test. In this case, the first detection information is determined as the second driving information of the surrounding unmanned vehicle. Thus, the comprehensiveness of the simulation data is ensured during the full analysis and solution of the problems related to the collaborative operation of unmanned vehicles.

[0086] In another embodiment of the present disclosure, in the above step S105, according to the output of the first algorithm, the input information of the subsequent first algorithm is adjusted in real time, including the following steps:

[0087] Step 1: Obtain the new first driving information of the unmanned vehicle included in the output of the current first algorithm;

[0088] Step 2: Use the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of the unmanned vehicle for sharing;

[0089] Step 3: And determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information.

[0090] In the embodiments of the present disclosure, the first algorithm may be the object to be tested. During the regression test stage of the first algorithm, the first algorithm may optimize the decision-making strategy of the unmanned vehicle, and the output of the first algorithm may be updated compared with the running record of the original log file. Therefore, it is necessary to adjust the input information of the subsequent first algorithm in real time according to the output of the first algorithm.

[0091] For the above step 1, the first algorithm is called during the playback process. The first algorithm may include various algorithms corresponding to each module in the log file. The call frame rate of various algorithms in the first algorithm is synchronized with the playback of the running record of the log file. After the first algorithm processes the input information, output data is obtained. During the playback process, the new first driving information of the unmanned vehicle included in the output of the current first algorithm is obtained. The new first driving information may include information such as the real-time position information, real-time vehicle speed, blind area, detected obstacle information, and driving trajectory planning of the current unmanned vehicle.

[0092] Regarding the above-mentioned step 2, use the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle for sharing. The first driving information shared to the shared data storage is used as the new first shared information. Therefore, the first shared information of the surrounding driverless vehicles obtained from the shared data storage may be inconsistent with the first detection information of the surrounding driverless vehicles extracted from the log file of this driverless vehicle.

[0093] Regarding the above-mentioned step 3, the first scenario information, the new first driving information, and the second driving information can be used as the input information for the subsequent first algorithm. Alternatively, according to the actual input parameter requirements of various algorithms in the first algorithm, after extracting the first scenario information, the new first driving information, and the second driving information, the input information for the subsequent first algorithm can be obtained. Using the new first driving information as one of the sources of the input information for the first algorithm can ensure the accuracy of the simulation test results.

[0094] In another embodiment of the present disclosure, after obtaining the new first driving information, it further includes:

[0095] Step 1: Compare the new first driving information with the first driving information recorded at the corresponding timestamp in the log file of this driverless vehicle;

[0096] Step 2: Use the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of this driverless vehicle for sharing;

[0097] In the above-mentioned step 3, determining the input information for the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information includes:

[0098] Step 3: In the case where the comparison results are inconsistent, use the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of this driverless vehicle for sharing; and determine the input information for the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information;

[0099] This method may further include:

[0100] Step 4: In the case where the comparison results are consistent, use the first driving information recorded at the corresponding timestamp in the log file of this driverless vehicle for sharing; and determine the input information for the subsequent first algorithm according to the first scenario information, the first driving information, and the second driving information.

[0101] In an embodiment of the present disclosure, after comparing the new first driving information with the first driving information recorded at the corresponding timestamp in the log file, the input information of the first algorithm is determined. For the above step 1, during the playback process, the new first driving information of the driverless vehicle included in the output of the current first algorithm is compared with the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle.

[0102] For the above step 2, the new first driving information is used to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle for sharing. After the first algorithm outputs the new first driving information, the new first driving information can be directly shared and shared to the shared data storage as the first shared information to improve the simulation efficiency.

[0103] For the above step 3, during the regression test stage of the first algorithm, the new first driving information may include the real-time position information, real-time vehicle speed, blind spot, detected obstacle information, and driving trajectory planning, etc. of the current driverless vehicle. In the case where the comparison results are inconsistent, the new first driving information is used to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle for sharing, and the first scenario information, the new first driving information, and the second driving information are used as the input information of the subsequent first algorithm, or after extracting the first scenario information, the new first driving information, and the second driving information according to the actual input parameter requirements of various algorithms in the first algorithm, the input information of the subsequent first algorithm is obtained.

[0104] For the above step 4, in the case where the comparison results are consistent, the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle is used for sharing to reduce the operation process and improve the simulation efficiency. And the first scenario information, the first driving information, and the second driving information are used as the input information of the subsequent first algorithm, or after extracting the first scenario information, the first driving information, and the second driving information according to the actual input parameter requirements of various algorithms in the first algorithm, the input information of the subsequent first algorithm is obtained. During the problem reproduction stage or regression test stage of the first algorithm, using the new first driving information as one of the sources of the input information of the first algorithm can ensure the accuracy of the simulation test results.

[0105] In another embodiment of the present disclosure, the following method is used to obtain the log files generated during the operation of multiple driverless vehicles participating in the simulation test:

[0106] Step 1, obtaining the log files generated during the operation of multiple driverless vehicles participating in the simulation test based on the method of real vehicle latch recording; and / or,

[0107] Step 2, obtaining the log files generated during the operation of multiple driverless vehicles participating in the simulation test based on the method of simulation recording.

[0108] In the embodiments of the present disclosure, the log file records key information such as the raw data, events that occurred, changes in the vehicle state, error messages, and relevant operations of algorithm modules during the operation of the driverless vehicle. The log files generated during the operation of multiple driverless vehicles participating in the simulation test can be obtained based on the method of real-vehicle latch recording or simulation recording.

[0109] Regarding the above step 1, recording the log file based on real-vehicle latch can be a technology for recording sensor data, vehicle state, environmental information, and system behavior generated during the operation of a driverless vehicle in a real operating environment. The latch mechanism can be triggered during critical events to save the complete data before and after the event, add an accurate timestamp to each operation record, and ensure the temporal synchronization of these operation record data. Use an efficient storage format, such as the pcap format, to save the log file and reduce the storage space occupied. Exemplarily, in an emergency braking scenario, when the driverless vehicle detects an obstacle ahead during driving and causes an emergency brake, the latch mechanism is triggered at this time, and the complete log file before and after the event is saved, such as the obstacle information collected by the sensor, the real-time position and vehicle speed information of the driverless vehicle, and the emergency braking decision information taken by the system when recognizing the obstacle. By playing back the log file, analyze the performance of the algorithm in the emergency braking scenario and optimize the algorithm.

[0110] Regarding the above step 2, the method based on simulation recording can provide the log files generated during the operation of the driverless vehicle in a simulation environment. For example, based on the WordSim simulation technology, a controllable and efficient simulation environment can be provided for the driverless vehicle. According to the cases of driverless vehicle collaborative operation problems, construct a simulation scenario file. For example, in the scenario of driverless vehicle collaborative operation in a mine, driverless vehicle a completes the loading task and drives away from the loading position, and driverless vehicles b and c enter the loading area and go to the loading position and the position to be loaded respectively. According to this example, the simulation platform can customize and edit the environmental information such as the loading position and the position to be loaded involved in this scenario, and simulate various sensors of the driverless vehicle, such as cameras, lidar, etc. Integrate algorithms such as decision-making and planning of the driverless vehicle into the simulation platform, and the operation process of the driverless vehicle in the scenario can be simulated. Obtaining the log files generated during the operation of multiple driverless vehicles based on the method of simulation recording can significantly reduce the dependence on real road scenarios, reduce costs, and improve the reliability of the algorithm.

[0111] In another embodiment of the present disclosure, the first algorithm includes a second algorithm corresponding to the operation record of the log file or a third algorithm updated based on the second algorithm.

[0112] During the simulation test, it is necessary to ensure the compatibility between the first algorithm and the original log file. The input information obtained by processing the log files generated during the operation of multiple unmanned vehicles participating in the simulation test should be recognizable and processable by the first algorithm. This requires the first algorithm to be compatible with the log file, and the log file generated after being processed by the first algorithm can still be used for subsequent simulation tests. Therefore, the first algorithm may include a second algorithm corresponding to the operation record of the log file or a third algorithm updated based on the second algorithm to ensure compatibility with the log file.

[0113] Exemplarily, as Figure 3 shown, in the problem reproduction stage, the second algorithm corresponds to the operation record of the log file. The log files generated during the operation of multiple unmanned vehicles participating in the simulation test, which are obtained, are logfile1, logfile12, and logfile3 respectively. After synchronous playback using flow control and logsim technology, the determined input information is input into the second algorithm to reproduce the problem. In the regression test stage, the third algorithm is updated based on the second algorithm. For example, optimizing the second algorithm for a large-scale unmanned driving cluster can ensure compatibility with logfile1, logfile12, and logfile3. After synchronous playback of logfile1, logfile12, and logfile3 using flow control and logsim technology, the determined input information is input into the third algorithm to conduct a regression test on the problem. That is to say, for the problems found in the problem reproduction stage, the second algorithm used in the problem reproduction stage can be upgraded and updated to obtain the third algorithm. Then, in the regression test stage, by synchronously playing back the log files generated during the operation of multiple unmanned vehicles participating in the simulation test again, the technical effects achieved by using the third algorithm in the same scenario can be determined.

[0114] Exemplarily, in the scenario of collaborative operation of unmanned vehicles in a mine, unmanned vehicle a completes the loading task and drives away from the loading position, while the other two unmanned vehicles b and c enter the loading area and head towards the loading position and the position to be loaded. Based on the WordSim simulation technology, a controllable and efficient simulation environment is provided to build the above scenario, and the log files generated during the operation of multiple unmanned vehicles participating in the simulation test are obtained. In the problem reproduction stage, by synchronously playing back the operation records of the log files corresponding to multiple unmanned vehicles respectively, and inputting the determined input information into the second algorithm, the result is as Figure 4The output information shown. The horizontal axis represents time, and the vertical axis represents the change in vehicle speed. The output information obtained during the problem reproduction stage is basically consistent with the running state of the driverless vehicle in the simulation scenario built based on the WordSim simulation technology. During the regression test stage, using the same log file, by synchronously playing back the running records of the log files corresponding to multiple driverless vehicles respectively, and inputting the determined input information into the third algorithm, the following is obtained Figure 5 The output information shown. The third algorithm can be updated based on the second algorithm to optimize the driving decision of the driverless vehicle. The horizontal axis represents time, and the vertical axis represents the change in vehicle speed. In the output information during the regression test stage, by comparing the information on the change in vehicle speed over time with the output information during the problem reproduction stage, it can be found that the driverless vehicle a shows different results during the time period when the speed is negative. Its physical meaning lies in the change in the reverse state of the vehicle. During the problem reproduction stage, the driverless vehicle a stopped midway during reverse. After being updated to the third algorithm, the driverless vehicle a completed a continuous reverse action during the regression test.

[0115] Based on the same general inventive concept, the embodiments of the present disclosure also provide a simulation test device and a driverless vehicle for a driverless vehicle. Since the principles of the problems solved by these devices and the driverless vehicle are similar to those of the aforementioned simulation test method for a driverless vehicle, the implementation of these devices and the driverless vehicle can refer to the implementation of the aforementioned method, and the repeated parts will not be elaborated.

[0116] The embodiments of the present disclosure provide a simulation test device for a driverless vehicle, as Figure 6 shown, including:

[0117] A log acquisition module 601, configured to acquire log files generated during the operation of multiple driverless vehicles participating in the simulation test;

[0118] An input information determination module 602, configured to synchronously play back the running records of the log files corresponding to the multiple driverless vehicles respectively, extract first scenario information, so as to simulate the historical running scenarios of the multiple driverless vehicles; and

[0119] During the playback process, for each driverless vehicle, according to the running records of the log file, the first driving information of the driverless vehicle is determined and shared in real time, and when there are surrounding driverless vehicles within the preset range of the driverless vehicle, according to the first shared information shared by the surrounding driverless vehicles in real time, the second driving information of the surrounding driverless vehicles is determined;

[0120] According to the first scenario information, the first driving information, and the second driving information, the input information of the first algorithm corresponding to the driverless vehicle is determined, and the input information is input into the first algorithm;

[0121] The output module 603 is configured to adjust the input information of the subsequent first algorithm in real time according to the output of the first algorithm until the playback is completed, so as to obtain the test result of the first algorithm.

[0122] In another embodiment of the present disclosure, the input information determination module 602 is configured to determine the playback frame rate according to the running record frame rates respectively corresponding to each module included in the log file;

[0123] Align the running records of each module included in the log files respectively corresponding to the multiple unmanned vehicles according to a preset time stamp, and control the synchronous playback of the playback frame rate to extract the first scenario information;

[0124] Simulate the historical running scenarios of the multiple unmanned vehicles according to the first scenario information.

[0125] In another embodiment of the present disclosure, when there are surrounding unmanned vehicles within the preset range of the unmanned vehicle, the input information determination module 602 is further configured to:

[0126] Determine the first detection information of the surrounding unmanned vehicle according to the running record of the log file; and

[0127] Determine whether the surrounding unmanned vehicle belongs to the multiple unmanned vehicles according to the identification information corresponding to the surrounding unmanned vehicle in the first detection information;

[0128] The input information determination module 602 is configured to, if it belongs, determine the first shared information shared in real time by the surrounding unmanned vehicle as the second driving information of the surrounding unmanned vehicle; or,

[0129] If it belongs, compare the first detection information with the first shared information shared in real time by the surrounding unmanned vehicle; in the case where the comparison result shows that the first detection information is inconsistent with the first shared information, determine the first shared information as the second driving information of the surrounding unmanned vehicle.

[0130] In another embodiment of the present disclosure, the input information determination module 602 is further configured to:

[0131] If it belongs and the comparison result is consistent, determine the first detection information as the second driving information of the surrounding unmanned vehicle; and / or,

[0132] If it does not belong, determine the first detection information as the second driving information of the surrounding unmanned vehicle.

[0133] In another embodiment of the present disclosure, the output module 603 is configured to

[0134] Obtain the new first driving information of the unmanned vehicle included in the output of the current first algorithm;

[0135] Share using the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and

[0136] Determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information.

[0137] In another embodiment of the present disclosure, after obtaining the new first driving information, the output module 603 is further configured to:

[0138] Compare the new first driving information with the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle;

[0139] Share using the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information, including:

[0140] In the case where the comparison results are inconsistent, share using the new first driving information to replace the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information;

[0141] The output module 603 is further configured to:

[0142] In the case where the comparison results are consistent, share the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the first driving information, and the second driving information.

[0143] In another embodiment of the present disclosure, the log acquisition module 601 is configured to acquire the log files generated during the operation of multiple driverless vehicles participating in the simulation test in the following manner:

[0144] Acquire the log files generated during the operation of multiple driverless vehicles participating in the simulation test based on the method of real vehicle latch recording; and / or,

[0145] Acquire the log files generated during the operation of multiple driverless vehicles participating in the simulation test based on the method of simulation recording.

[0146] In another embodiment of the present disclosure, the first algorithm includes a second algorithm corresponding to the operation record of the log file or a third algorithm updated based on the second algorithm.

[0147] An embodiment of the present disclosure provides an autonomous vehicle, including: the simulation test device of the autonomous vehicle as described in any of the above embodiments.

[0148] From the description of the above embodiments, those skilled in the art can clearly understand that the embodiments of the present disclosure can be implemented by hardware, or can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solutions of the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments of the present disclosure.

[0149] Those skilled in the art can understand that the drawings are only schematic diagrams of a preferred embodiment, and the modules or processes in the drawings are not necessarily essential for implementing the present disclosure.

[0150] Those skilled in the art can understand that the modules in the device in the embodiment can be distributed in the device in the embodiment according to the description of the embodiment, or can be changed accordingly and located in one or more devices different from this embodiment. The modules of the above embodiments can be combined into one module, or can be further split into multiple sub-modules.

[0151] The serial numbers of the above embodiments of the present disclosure are only for description and do not represent the advantages and disadvantages of the embodiments.

[0152] Obviously, those skilled in the art can make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalent technologies, the present disclosure also intends to include these modifications and variations.

Claims

1. A simulation test method for an unmanned vehicle, characterized in that Including: Obtaining log files generated during the operation of multiple autonomous vehicles participating in simulation tests; Synchronously playing back the operation records of the log files respectively corresponding to the multiple autonomous vehicles, and extracting first scenario information to simulate the historical operation scenarios of the multiple autonomous vehicles; During the playback process, for each autonomous vehicle, according to the operation records of the log file, determining and sharing the first driving information of the autonomous vehicle in real time, and when there are surrounding autonomous vehicles within the preset range of the autonomous vehicle, determining the second driving information of the surrounding autonomous vehicles according to the first shared information shared by the surrounding autonomous vehicles in real time; Determining the input information of the first algorithm corresponding to the autonomous vehicle according to the first scenario information, the first driving information, and the second driving information, and inputting the input information into the first algorithm; According to the output of the first algorithm, adjusting the input information of the subsequent first algorithm in real time until the playback is completed to obtain the test result of the first algorithm.

2. The method according to claim 1, characterized in that, The synchronously playing back the operation records of the log files respectively corresponding to the multiple autonomous vehicles, and extracting first scenario information to simulate the historical operation scenarios of the multiple autonomous vehicles includes: Determining the playback frame rate according to the operation record frame rates respectively corresponding to the various modules included in the log file; Aligning the operation records of the various modules included in the log files respectively corresponding to the multiple autonomous vehicles according to a preset time stamp, and controlling the playback frame rate for synchronous playback to extract first scenario information; Simulating the historical operation scenarios of the multiple autonomous vehicles according to the first scenario information.

3. The method according to claim 1, wherein When there are surrounding autonomous vehicles within the preset range of the autonomous vehicle, it further includes: Determining the first detection information of the surrounding autonomous vehicles according to the operation records of the log file; and Determining whether the surrounding autonomous vehicles belong to the multiple autonomous vehicles according to the identification information corresponding to the surrounding autonomous vehicles in the first detection information; The determining the second driving information of the surrounding autonomous vehicles according to the first shared information shared by the surrounding autonomous vehicles in real time includes: If it belongs, determining the first shared information shared by the surrounding autonomous vehicle in real time as the second driving information of the surrounding autonomous vehicle; or, If it belongs, comparing the first detection information with the first shared information shared by the surrounding autonomous vehicle in real time; when the comparison result shows that the first detection information is inconsistent with the first shared information, determining the first shared information as the second driving information of the surrounding autonomous vehicle.

4. The method according to claim 3, wherein It further includes: If it belongs and the comparison result is consistent, determining the first detection information as the second driving information of the surrounding autonomous vehicle; And / or, If it does not belong, determining the first detection information as the second driving information of the surrounding autonomous vehicle.

5. The method according to claim 1, wherein The adjusting the input information of the subsequent first algorithm in real time according to the output of the first algorithm includes: Obtaining the new first driving information of the autonomous vehicle included in the output of the current first algorithm; Using the new first driving information to replace the first driving information recorded at the corresponding time stamp in the log file of the autonomous vehicle for sharing; and Determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information.

6. The method according to claim 5, characterized in that, After obtaining the new first driving information, it further includes: Compare the new first driving information with the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; Share the new first driving information instead of the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information, including: In the case where the comparison result is inconsistent, share the new first driving information instead of the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the new first driving information, and the second driving information; The method further includes: In the case where the comparison result is consistent, share the first driving information recorded at the corresponding timestamp in the log file of the driverless vehicle; and determine the input information of the subsequent first algorithm according to the first scenario information, the first driving information, and the second driving information.

7. The method according to claim 1, wherein Obtain the log files generated during the operation of multiple driverless vehicles participating in the simulation test in the following manner: Obtain the log files generated during the operation of multiple driverless vehicles participating in the simulation test based on the method of real vehicle latch recording; and / or Obtain the log files generated during the operation of multiple driverless vehicles participating in the simulation test based on the method of simulation recording.

8. The method according to claim 1, wherein The first algorithm includes a second algorithm corresponding to the operation record of the log file or a third algorithm updated based on the second algorithm.

9. A simulation test device for an autonomous vehicle, characterized in that, It includes: A log acquisition module for obtaining the log files generated during the operation of multiple driverless vehicles participating in the simulation test; An input information determination module for synchronously playing back the operation records of the log files corresponding to the multiple driverless vehicles respectively, extracting the first scenario information to simulate the historical operation scenarios of the multiple driverless vehicles; and During the playback process, for each driverless vehicle, determine and share the first driving information of the driverless vehicle in real time according to the operation record of the log file, and in the case where there are surrounding driverless vehicles within the preset range of the driverless vehicle, determine the second driving information of the surrounding driverless vehicles according to the first shared information shared by the surrounding driverless vehicles in real time; Determine the input information of the first algorithm corresponding to the driverless vehicle according to the first scenario information, the first driving information, and the second driving information, and input the input information into the first algorithm; An output module for adjusting the input information of the subsequent first algorithm in real time according to the output of the first algorithm until the playback is completed to obtain the test result of the first algorithm.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, it executes the steps of the method for simulating and testing a driverless vehicle according to any one of claims 1 to 8.

Images