A test method for improving simulation test efficiency in autonomous driving requirement verification

Abstract

The present invention discloses a test method for improving the efficiency of simulation testing in automatic driving demand verification, and the steps include: configuring a host, a server and multiple slaves; configuring a list of test cases to be executed on the host according to a proposed test plan, and the host sends the list of test cases to the server; the host determines whether the slave is available according to the slave status log recorded on the server, and if so, sets it as available in the slave list; otherwise, sets it as unavailable in the slave list; counting available slaves to form a node list; the host determines whether the node is idle according to the slave status log recorded on the server, and if so, the host allocates a single test task to the slave corresponding to the idle node, and the slave receives the test task and executes the test task, and saves the simulation test result to the server. This method can speed up the test process of intelligent driving virtual testing, improve the efficiency of model simulation testing, shorten the test cycle and reduce the test cost.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous driving simulation testing, and in particular to a testing method for improving the efficiency of simulation testing in autonomous driving requirement verification. Background Art

[0002] At present, the intelligent driving technology of intelligent connected vehicles involves a large number of tests in the research and development process, including virtual tests, closed-field tests, and open road tests. In the early stage of research and development, virtual simulation testing is the main testing method for algorithm development and function verification. Simulation testing involves virtual scenes, including traffic participants, static road models, environmental models, etc. The virtual scenes have very high hardware requirements during the virtual testing process, and a single working condition and use case test will take a certain amount of time; at the same time, the use of high-precision vehicle dynamics models means increasing the simulation frequency by up to dozens of times, which brings huge computing pressure to the simulation test platform. When there are many test cases, the efficiency of virtual testing will be greatly challenged, and even affect the development progress. Summary of the invention

[0003] The purpose of the present invention is to provide a testing method for improving the efficiency of simulation testing in autonomous driving demand verification, which can speed up the testing process of intelligent driving virtual testing, improve the efficiency of model simulation testing, shorten the testing cycle and reduce the testing cost.

[0004] To achieve the above object, the present invention provides a test method for improving the efficiency of simulation testing in autonomous driving requirement verification, the steps comprising:

[0005] (S1) configuring a host for participating in simulation testing, a server for storing data and distributing test tasks, and multiple slave machines for simulation test calculations;

[0006] (S2) configuring a list of test cases to be executed on the host according to the proposed test plan, and the host sends the list of test cases to the server;

[0007] (S3) The host determines whether the slave is available according to the slave status log recorded on the server. If so, it is set as available in the slave list; otherwise, it is set as unavailable in the slave list;

[0008] (S4) Counting available slaves to form a node list;

[0009] (S5) The host determines whether the node is idle according to the slave status log recorded on the server. If so, the host assigns a single test task to the slave corresponding to the idle node. The slave receives the test task and executes it, saves the simulation test result to the server, and goes to execute steps (S6) and (S7); otherwise, returns to step (S3);

[0010] (S6) The host detects whether all test tasks have been completed. If so, it goes to step (S10); otherwise, it returns to step (S3);

[0011] (S7) The slave reads the execution configuration file in the server through the background program;

[0012] (S8) The slave determines whether the master has assigned a new test task. If so, it proceeds to step (S9); otherwise, it returns to step (S7);

[0013] (S9) The slave receives the test task and executes the test task, saves the simulation test result in the server, and goes to the execution step (S8);

[0014] (S10) Send a command to shut down all slaves, the host reads the test results on the data server, automatically generates a test report, and the host process ends.

[0015] Further, the slave receives a new test task and executes the test task, and specifically performs the following steps: the slave automatically generates a test model after receiving the test task, and the slave executes a simulation test program.

[0016] Furthermore, the proposed test plan includes writing test cases, test step documents and test script documents, setting up simulation test scenarios and simulation test platforms, configuring simulation test environments and installing simulation test software.

[0017] Furthermore, the simulation test platform includes a high-precision vehicle dynamics simulation test platform and a simulation test platform without vehicle dynamics, and the simulation test scenarios include static roads, dynamic traffic participants, operation of the driver model and jump logic of the vehicle information status.

[0018] Furthermore, the slave machine automatically generates a test model after receiving the test task, and specifically performs the following steps: after receiving the test task, the slave machine determines whether the test case requires a high-precision dynamic model. If so, a simulation platform for automatically generating a high-precision dynamic model is selected; otherwise, a simulation platform without a dynamic model is selected.

[0019] Compared with the prior art, the present invention has the following advantages:

[0020] The test method for improving the efficiency of simulation testing in the verification of autonomous driving requirements of the present invention accelerates the test process of intelligent driving virtual testing, improves the efficiency of model simulation testing, shortens the test cycle and reduces the test cost by summarizing and analyzing system requirements and distributing simulation tasks during the development of a limited autonomous driving system; idle computers in the office can be used as distributed computing nodes, and existing resources can be used to improve the efficiency of simulation testing. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 The present invention is a flow chart of a test method for improving the efficiency of simulation testing in autonomous driving requirement verification. DETAILED DESCRIPTION

[0022] The specific implementation of the present invention will be further described below in conjunction with the accompanying drawings.

[0023] See also Figure 1 As shown, this embodiment discloses a test method for improving the efficiency of simulation testing in automatic driving requirement verification, and the steps include:

[0024] (S1) configuring a host for participating in simulation testing, a server for storing data and distributing test tasks, and multiple slave machines for simulation test calculations;

[0025] (S2) A list of test cases to be executed is configured on the host according to the proposed test plan, and the host sends the list of test cases to the server; the test leader needs to prepare a test plan in advance before the project team releases the official version of the software, and after the software is officially released, a list of test cases is generated according to the list of test cases in the test plan.

[0026] (S3) The host determines whether the slave is available according to the slave status log recorded on the server. If so, it is set as available in the slave list; otherwise, it is set as unavailable in the slave list;

[0027] (S4) Counting available slaves to form a node list;

[0028] (S5) The host determines whether the node is idle according to the slave status log recorded on the server. If so, the host assigns a single test task to the slave corresponding to the idle node. The slave receives the test task and executes it, saves the simulation test result to the server, and goes to execute steps (S6) and (S7); otherwise, returns to step (S3);

[0029] (S6) The host detects whether all test tasks have been completed. If so, it goes to step (S10); otherwise, it returns to step (S3);

[0030] (S7) The slave reads the execution configuration file in the server through the background program;

[0031] (S8) The slave determines whether the host has assigned a new test task. If so, it goes to execution step (S9); otherwise, it returns to execution step (S7). Specifically, the slave reads the execution configuration file in the server through the background application to identify whether the host has assigned a test task.

[0032] (S9) The slave receives the test task and executes the test task, saves the simulation test result in the server, and goes to the execution step (S8);

[0033] (S10) Send instructions to shut down all slaves, the host reads the test results on the data server, automatically generates a test report, and the host process ends. It is necessary to have a high-performance computer as the host, a server, and multiple slaves as test nodes. The host distributes test tasks to the server, the slave reads the server test task file to obtain the test task, the slave executes the test case and uploads the test results to the server, and the host generates a test report based on the test results. The method of the present invention can connect multiple computers in a unified local area network through a server as a medium for information transmission, thereby achieving the purpose of linearly improving the efficiency of simulation testing.

[0034] In this embodiment, the slave receives a new test task and executes the test task, and specifically performs the following steps: the slave automatically generates a test model after receiving the test task, and the slave executes a simulation test program.

[0035] In this embodiment, the proposed test plan includes writing test cases, test step documents and test script documents, building simulation test scenarios and simulation test platforms, configuring simulation test environments and installing simulation test software.

[0036] In this embodiment, the simulation test platform includes a high-precision vehicle dynamics simulation test platform and a simulation test platform without vehicle dynamics, and the simulation test scenarios include static roads, dynamic traffic participants, operation of the driver model, and jump logic of the vehicle information status.

[0037] In this embodiment, the slave automatically generates a test model after receiving the test task, and specifically performs the following steps:

[0038] After receiving the test task, the slave determines whether the test case requires a high-precision dynamics model. If so, a simulation platform with a high-precision dynamics model is selected and automatically generated; otherwise, a simulation platform without a dynamics model is selected. For system requirements that only involve system state jumps or target selection but not system closed-loop control, a simulation platform without a dynamics model is used. If it is necessary to verify the closed-loop control or if this has a strong coupling relationship with the vehicle's kinematic parameters, a simulation test platform with high-precision vehicle dynamics is used. By drawing a trajectory for the main vehicle equipped with an autonomous driving system in the simulator, it is determined whether the actual performance of the state quantity is consistent with the expected performance, thereby achieving the purpose of simulation testing.

[0039] The test method for improving the efficiency of simulation testing in the verification of autonomous driving requirements of the present invention accelerates the test process of intelligent driving virtual testing, improves the efficiency of model simulation testing, shortens the test cycle and reduces the test cost by summarizing and analyzing system requirements and distributing simulation tasks in the development process of limited autonomous driving system. Idle computers in the office can be used as distributed computing nodes, and existing resources can be used to improve the efficiency of simulation testing.

[0040] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A test method for improving simulation test efficiency in autonomous driving requirements verification, It is characterized in that the steps include: (S1) configuring a host for participating in simulation testing, a server for storing data and distributing test tasks, and multiple slave machines for simulation test calculations; (S2) configuring a list of test cases to be executed on the host according to the proposed test plan, and the host sends the list of test cases to the server; (S3) The host determines whether the slave is available according to the slave status log recorded on the server. If so, it is set as available in the slave list; otherwise, it is set as unavailable in the slave list; (S4) Counting available slaves to form a node list; (S5) The host determines whether the node is idle according to the slave status log recorded on the server. If so, the host assigns a single test task to the slave corresponding to the idle node. The slave receives the test task and executes it, saves the simulation test result to the server, and goes to execute steps (S6) and (S7); otherwise, returns to step (S3); (S6) The host detects whether all test tasks have been completed. If so, it goes to step (S10); otherwise, it returns to step (S3); (S7) The slave reads the execution configuration file in the server through the background program; (S8) The slave determines whether the master has assigned a new test task. If so, it proceeds to step (S9); otherwise, it returns to step (S7); (S9) The slave receives the test task and executes the test task, saves the simulation test result in the server, and goes to the execution step (S8); (S10) Send a command to shut down all slaves, the host reads the test results on the data server, automatically generates a test report, and the host process ends.

2. The test method for improving simulation test efficiency in autonomous driving requirement verification according to claim 1, It is characterized in that The slave receives a new test task and executes the test task, specifically performing the following steps: After receiving the test task, the slave automatically generates a test model and executes the simulation test program.

3. The test method for improving simulation test efficiency in autonomous driving requirement verification according to claim 2, It is characterized in that The proposed test plan includes writing test cases, test step documents and test script documents, setting up simulation test scenarios and simulation test platforms, configuring simulation test environments and installing simulation test software.

4. The test method for improving simulation test efficiency in autonomous driving requirement verification according to claim 3, It is characterized in that The simulation test platform includes a high-precision vehicle dynamics simulation test platform and a simulation test platform without vehicle dynamics. The simulation test scenarios include static roads, dynamic traffic participants, driver model operations, and vehicle information status jump logic.

5. The test method for improving simulation test efficiency in autonomous driving requirement verification according to claim 4, It is characterized in that After receiving the test task, the slave automatically generates a test model and specifically performs the following steps: After receiving the test task, the slave determines whether the test case requires a high-precision dynamics model. If so, a simulation platform for the high-precision dynamics model is selected and automatically generated; otherwise, a simulation platform without a dynamics model is selected.

Images