Verification method, device and equipment for functional safety of autonomous driving simulation test

Abstract

This application relates to a method, apparatus, and device for verifying the functional safety of autonomous driving simulation testing. The method includes: establishing a scenario library for autonomous driving simulation testing, automatically generating logical scenarios corresponding to the scenario library based on complete parameters in a preset template; determining the design and operation domain (ODD) of the autonomous driving simulation test based on the logical scenarios based on a preset static map library, preset evaluation indicators, and a preset measurable scenario description language; generalizing the logical scenarios based on the autonomous driving simulation test ODD to obtain multiple target scenarios, and simulating the algorithm to be verified in the vehicle under test in each target scenario to obtain the safety results of the autonomous driving simulation test function. This method solves the problems of long time cycles, high investment costs, and low development efficiency in real-vehicle testing of autonomous driving systems, as well as the inefficiency of scenario and simulation scenario generation methods in related technologies, and can realize the vision of high efficiency, high quality, and low cost in autonomous driving simulation testing.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving simulation technology, and in particular to a method, device, and equipment for verifying the functional safety of autonomous driving simulation testing. Background Art

[0002] Autonomous driving technology not only improves traffic safety and alleviates traffic congestion, but also increases road efficiency. Currently, autonomous driving technology is maturing in terms of perception, decision-making, and control. However, safety concerns are hindering further development of existing technologies, making it particularly important to test the safety of autonomous driving systems.

[0003] In the related art, the safety of an autonomous driving system is usually tested based on simulation testing or real vehicle testing, or simulation testing is performed based on an ADAS (Advanced Driving Assistance System).

[0004] However, real-vehicle testing has a long cycle, high investment costs, and low development efficiency. The testing methods in related technologies are based on the ADAS (Advanced Driving Assistance System) system, but faced with complex known and unknown real-world environments, the scene and simulation scene generation methods are inefficient. In terms of scene generalization, effective test scenarios cannot be automatically generated, and the monitored vehicle behavior parameter data cannot be automatically analyzed, which urgently needs to be solved. Summary of the Invention

[0005] The present application provides a method, device and equipment for verifying the functional safety of autonomous driving simulation tests to solve the problems of long time cycle, high investment cost and low development efficiency in actual vehicle testing of autonomous driving systems, low efficiency of scene and simulation scene generation methods in related technologies, and inability to automatically generate effective test scenes in terms of scene generalization, and inability to automatically analyze the monitored vehicle behavior parameter data. The application can realize the vision of high efficiency, high quality and low cost of autonomous driving simulation testing, and has the advantages of scene specification, automated testing and big data analysis.

[0006] The first aspect of the present application provides a method for verifying the safety of an autonomous driving simulation test function, comprising the following steps: establishing a scenario library for the autonomous driving simulation test, and automatically generating a logical scenario corresponding to the scenario library based on complete parameters in a preset template; determining an operational design domain (ODD) (designed operational area) of the autonomous driving simulation test according to the logical scenario based on a preset static map library, preset evaluation indicators, and a preset measurable scenario description language; generalizing the logical scenario based on the ODD of the autonomous driving simulation test to obtain multiple target scenarios, and simulating and testing the algorithm to be verified of the vehicle to be tested in each target scenario to obtain a safety result of the autonomous driving simulation test function.

[0007] The above-mentioned technical means can solve the problems of long time cycle, high investment cost and low development efficiency in actual vehicle testing of autonomous driving systems, low efficiency of scene and simulation scene generation methods in related technologies, and inability to automatically generate effective test scenes in scene generalization, and the inability to automatically analyze the monitored vehicle behavior parameter data. It can realize the vision of high efficiency, high quality and low cost of autonomous driving simulation testing, and has the advantages of scene specification, automated testing and big data analysis.

[0008] Furthermore, after obtaining the safety result of the autonomous driving simulation test function, it also includes: extracting the target scenario where the simulation test fails from the safety result; obtaining the motion state information of the vehicle to be tested under the target scenario where the simulation test fails; and generating the reason for the failure of the simulation test based on the motion state information of the vehicle to be tested under the target scenario where the simulation test fails.

[0009] According to the above technical means, the corresponding specific scenario is found through the corresponding random seed, so as to analyze the vehicle motion state in the specific scenario, which can further improve the tested autonomous driving system.

[0010] Furthermore, after simulating and testing the algorithm to be verified of the vehicle to be tested in each target scene, it also includes: obtaining a random seed corresponding to the current target scene; and storing the motion state information of the current vehicle in the current target scene generated by the random seed corresponding to the current target scene to a target position.

[0011] According to the above technical means, the data generated by the target scenario generated by the random seed is stored in the target location, which is convenient for extracting the target scenarios of success and failure of the simulation test.

[0012] Furthermore, the automatic generation of a logical scene corresponding to the scene library based on the complete parameters in the preset template includes: obtaining the Ego (custom mode) vehicle speed, the non-player character NPC (Non-Player Character) vehicle speed, the longitudinal distance between the Ego speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scene, road geometry information, and road facility information; automatically generating a logical scene corresponding to the scene library based on at least one parameter space among the Ego speed, the NPC vehicle speed, the longitudinal distance between the Ego speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scene, the road geometry information, and the road facility information.

[0013] Based on the above technical means, through the method of multi-dimensional parameter space, the Ego vehicle speed, NPC vehicle speed, the longitudinal distance between the Ego vehicle speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scene, the road geometry information, and the road facility information in real life are generalized from logical scenes to specific scenes.

[0014] Furthermore, the scenario library includes at least one of a highway scenario library, an accident scenario library, an expert scenario library and an extreme weather scenario library.

[0015] According to the above technical means, autonomous driving simulation tests are carried out through multiple scenario libraries, and the safety results of autonomous driving simulation tests are more accurate.

[0016] The second aspect of the present application provides a device for verifying the safety of an autonomous driving simulation test function, including: an establishment module for establishing a scenario library for autonomous driving simulation testing, and automatically generating a logical scenario corresponding to the scenario library based on complete parameters in a preset template; a determination module for determining the design operation domain (ODD) of the autonomous driving simulation test according to the logical scenario based on a preset static map library, preset evaluation indicators and a preset measurable scenario description language; a testing module for generalizing the logical scenario based on the ODD of the autonomous driving simulation test to obtain multiple target scenarios, and simulating and testing the algorithm to be verified of the vehicle to be tested in each target scenario to obtain the safety result of the autonomous driving simulation test function.

[0017] Furthermore, after obtaining the safety result of the autonomous driving simulation test function, it also includes: an extraction module for extracting the target scenario where the simulation test fails from the safety result; an acquisition module for obtaining the motion state information of the vehicle to be tested under the target scenario where the simulation test fails; and a generation module for generating the reason for the failure of the simulation test based on the motion state information of the vehicle to be tested under the target scenario where the simulation test fails.

[0018] Furthermore, the test module is specifically used to: obtain a random seed corresponding to the current target scene;

[0019] The motion state information of the current vehicle in the current target scene generated by the random seed corresponding to the current target scene is stored in the target position.

[0020] Furthermore, the automatic generation of logical scenarios corresponding to the scenario library based on the complete parameters in the preset template includes: obtaining the Ego vehicle speed, the non-player character NPC vehicle speed, the longitudinal distance between the Ego vehicle speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scenario, road geometry information, and road facility information; automatically generating the logical scenarios corresponding to the scenario library based on at least one parameter space among the Ego vehicle speed, the NPC vehicle speed, the longitudinal distance between the Ego vehicle speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scenario, the road geometry information, and the road facility information.

[0021] Furthermore, the scenario library includes at least one of a highway scenario library, an accident scenario library, an expert scenario library and an extreme weather scenario library.

[0022] The third aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement a method for verifying the functional safety of an autonomous driving simulation test as described in the above embodiment.

[0023] The fourth aspect of the present application provides a computer-readable storage medium on which a computer program is stored. The program is executed by a processor to implement the method for verifying the functional safety of the autonomous driving simulation test as described in the above embodiment.

[0024] Thus, by establishing a scenario library for autonomous driving simulation testing, logical scenarios corresponding to the scenario library are automatically generated based on the complete parameters in a preset template. Based on a preset static map library, preset evaluation indicators, and a preset measurable scenario description language, the design and operation domain (ODD) for the autonomous driving simulation test is determined based on the logical scenarios. Based on the ODD for the autonomous driving simulation test, the logical scenarios are generalized to obtain multiple target scenarios. The algorithm to be verified for the vehicle under test is simulated and tested in each target scenario to obtain safety results for the autonomous driving simulation test function. This solves the long time, high investment cost, and low development efficiency of real-vehicle autonomous driving system testing, the inefficiency of scenario and simulation scenario generation methods in related technologies, the inability to automatically generate effective test scenarios when generalizing scenarios, and the inability to automatically analyze monitored vehicle behavior parameter data. This system can achieve the vision of high-efficiency, high-quality, and low-cost autonomous driving simulation testing, while also offering advantages such as scenario standardization, automated testing, and big data analysis.

[0025] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0027] Figure 1 This is a flowchart of a method for verifying the safety of an autonomous driving simulation test function according to an embodiment of the present application;

[0028] Figure 2 This is a schematic diagram of the overall framework flow of autonomous driving simulation testing according to one embodiment of the present application;

[0029] Figure 3 A schematic diagram of a measurable scenario description language according to an embodiment of the present application;

[0030] Figure 4 1 is a block diagram of important components of an autonomous driving system according to one embodiment of the present application;

[0031] Figure 5 Schematic diagram of a block diagram of a device for verifying functional safety of an autonomous driving simulation test according to an embodiment of the present application;

[0032] Figure 6 Schematic diagram of the structure of an electronic device according to an embodiment of the present application.

[0033] Explanation of the reference numerals: 10 - verification device for functional safety of autonomous driving simulation test, 100 - establishment module, 200 - determination module, 300 - test module, 601 - memory, 602 - processor, 603 - communication interface. DETAILED DESCRIPTION

[0034] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0035] The following describes the method, device and equipment for verifying the functional safety of autonomous driving simulation testing of an embodiment of the present application with reference to the accompanying drawings. In order to solve the problems mentioned in the above background technology of long time period, high investment cost and low development efficiency of the autonomous driving system tested on a real vehicle, as well as the low efficiency of the scene and simulation scene generation methods in the related technology, the present application provides a method for verifying the functional safety of autonomous driving simulation testing. In this method, by establishing a scene library for autonomous driving simulation testing, a logical scene corresponding to the scene library is automatically generated based on the complete parameters in the preset template. Based on the preset static map library, preset evaluation indicators and preset measurable scene description language, the design operation domain (ODD) of the autonomous driving simulation test is determined according to the logical scene. Based on the ODD of the autonomous driving simulation test, the logical scene is generalized to obtain multiple target scenes, and the algorithm to be verified of the vehicle to be tested is simulated and tested in each target scene to obtain the safety result of the autonomous driving simulation test function. This solves the problems of long time cycle, high investment cost and low development efficiency in actual vehicle testing of autonomous driving systems, inefficient scene and simulation scene generation methods in related technologies, inability to automatically generate effective test scenes in scene generalization, and inability to automatically analyze monitored vehicle behavior parameter data. It can realize the vision of high efficiency, high quality and low cost of autonomous driving simulation testing, and has the advantages of scene specification, automated testing and big data analysis.

[0036] Specifically, Figure 1 A flowchart of a method for verifying the safety of an autonomous driving simulation test function provided in an embodiment of the present application.

[0037] like Figure 1 As shown, the verification method for the functional safety of the autonomous driving simulation test includes the following steps:

[0038] In step S101, a scenario library for autonomous driving simulation testing is established, and a logical scenario corresponding to the scenario library is automatically generated based on the complete parameters in the preset template.

[0039] Furthermore, in some embodiments, the scenario library includes at least one of a highway scenario library, an accident scenario library, an expert scenario library, and an extreme weather scenario library.

[0040] It should be understood that if Figure 2 As shown, the scenario library can come from the highway scenario library, accident scenario library, expert scenario library, extreme weather scenario library, etc. The establishment of the scenario library can come from real vehicle data and also from simulation test data.

[0041] Furthermore, in some embodiments, a logical scene corresponding to the scene library is automatically generated based on the complete parameters in the preset template, including: obtaining the Ego vehicle speed, the non-player character NPC vehicle speed, the longitudinal distance between the Ego vehicle speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scene, road geometry information, and road facility information; and automatically generating a logical scene corresponding to the scene library based on at least one parameter space among the Ego vehicle speed, the NPC vehicle speed, the longitudinal distance between the Ego vehicle speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scene, the road geometry information, and the road facility information.

[0042] Specifically, after having a complete scenario library, the embodiment of the present application can automatically generate corresponding logical scenarios through the provided template tool based on the complete parameters required to be provided in the template. The logical scenario is between the functional scenario and the specific scenario, and its abstraction gradually decreases, while the number of opposite scenarios increases. The logical scenario contains some multi-dimensional parameter spaces, including the Ego vehicle speed, the NPC vehicle speed, the longitudinal relative distance between the Ego and the NPC, the NPC vehicle lane, the duration of each stage in the logical scenario, road geometry information, road facility information, etc. These parameters are not specific values, but a parameter space method to construct the logical scenario.

[0043] It should be noted that the complete parameters of the template tool can be parameters pre-set by the user, parameters obtained through a limited number of experiments, or parameters obtained through a limited number of computer simulations, and no specific limitation is made here.

[0044] In step S102, based on a preset static map library, preset evaluation indicators and a preset measurable scenario description language, a design operation domain (ODD) of the autonomous driving simulation test is determined according to a logical scenario.

[0045] Among them, the preset evaluation indicators may include statistical evaluation indicators, scenario-specific evaluation indicators, and multi-dimensional general evaluation indicators.

[0046] It should be understood that if Figure 3As shown in the figure, in the autonomous driving simulation test system, scenarios are usually divided into functional scenarios, logical scenarios, and specific scenarios. Among them, the logical scenario is based on a parameter space, and each parameter has a specific range. In this way, static and dynamic scenarios are described. It includes reusable basic scenarios and dynamic scenarios. When synthesizing more complex advanced scenarios, third-party module libraries can be imported to form more complex scenarios, increasing the reusability of the scenarios. At the same time, the scenarios also need to have scene triggers, static maps based on the OpenDrive format, and corresponding evaluation indicators. Therefore, a measurable scenario description language based on Python is used as a link to combine various important sub-modules. To trigger a scenario, a main function must be included to enable the entire code to run properly. For example, the scenario trigger code "extend top.main" serves as the entry point for the entire automation program. Furthermore, to conduct autonomous driving simulation tests under urban conditions, a corresponding static map is selected from the static map library, such as "import. / hooder.xodr." Furthermore, the simulation test functions involve some basic evaluation metrics, such as Ego speed and distance to the left lane line. Furthermore, specific parameters corresponding to specific functional tests include the time to customer (TTC) and the cut-in direction of the preceding vehicle. Furthermore, extensive intelligent driving simulation tests are conducted to obtain the design operating domain (ODD). Therefore, the ODD range can be determined statistically by observing specific parameters, such as "extenddut.two_cut_in:cover(car1.speed)." The design operating domain (ODD) refers to the operating conditions set for the autonomous driving system functions, including but not limited to environmental, geographical, and time-of-day restrictions, traffic flow, and road characteristics. Only when all conditions are met can the autonomous driving be guaranteed to operate normally.

[0047] In step S103, based on the ODD of the autonomous driving simulation test, the logical scenario is generalized to obtain multiple target scenarios, and the algorithm to be verified of the vehicle to be tested is simulated and tested in each target scenario to obtain the safety result of the autonomous driving simulation test function.

[0048] Specifically, the ODD scenario generalization tool based on autonomous driving simulation testing uses the application security testing software Foretify. Its generalization principle is based on relevant statistical theories. Taking into account the vehicle dynamics model, the logical scenario will be generalized to generate specific scenarios. The global and local trajectories of the NPC vehicle's specific movement after generalization are composed of a series of control points. Among them, each logical scenario can be generalized to generate at least one thousand different specific scenarios. Each specific scenario will correspond to a different random seed, and the correspondence between the random seed and the specific scenario is robust.

[0049] This simulation tool is based on the OpenX 2.0 standard and can be co-simulated with dynamics simulation software such as CarSim (vehicle dynamics simulation software). It can also be connected to an external DSP (Digital Signal Processing), that is, an external SIL (Software in the loop) or MIL (Model in the Loop) model, to simulate and test the algorithm under test that controls the vehicle. At the same time, during the simulation test, the relevant motion status information of the vehicle can be printed, and the entire simulation process can be played back in animation or single-step simulation debugging.

[0050] Furthermore, in some embodiments, after simulating and testing the algorithm to be verified of the vehicle to be tested in each target scene, it also includes: obtaining a random seed corresponding to the current target scene; and storing the motion state information of the current vehicle in the current target scene generated by the random seed corresponding to the current target scene to the target position.

[0051] It is understandable that, based on the simulation tool during the simulation process, the embodiment of the present application will store the motion state information data of the vehicle generated for each specific scenario generated by the random seed in the metrics_data.json file.

[0052] Furthermore, in some embodiments, after obtaining the safety results of the autonomous driving simulation test function, it also includes: extracting the target scenario where the simulation test fails from the safety results; obtaining the motion state information of the vehicle to be tested under the target scenario where the simulation test fails; and generating the reason for the failure of the simulation test based on the motion state information of the vehicle to be tested under the target scenario where the simulation test fails.

[0053] Specifically, through big data analysis tools, the parameter data of the evaluation indicators of all generalized specific scenarios under a certain logical scenario corresponding to the functional safety of the autonomous driving simulation test will be loaded. The panel contains scenario coverage and KPI (Key Performance Indicator) indicators. There is a direct connection between scenario coverage and KPI indicators. The mean square sum of the coverage of each evaluation indicator is the scenario coverage. In a large number of generalized specific scenarios, due to the deficiencies and defects of the autonomous driving system, there will also be specific scenarios of failure. At this time, the corresponding specific scenario can be found through the corresponding random seed, so as to analyze the vehicle motion state information in the specific scenario, find the real cause of the failure, and further improve the autonomous driving system of the object under test.

[0054] Among them, Figure 4 As shown in the figure, the autonomous driving system uses algorithms in four aspects: perception, decision-making, planning and control. The perception algorithm is based on a monocular camera, and uses the target detection algorithm yolov5 and the multi-target tracking algorithm deepsort to identify, classify and track the target vehicle. At the same time, the Zhang Zhengyou calibration method is used to calibrate the internal and external parameters of the monocular camera, and then complete the distance measurement algorithm for the target vehicle. Under the premise of perceiving the surrounding environment, after obtaining the motion state information of the surrounding vehicle and the surrounding vehicles, including speed, acceleration, acceleration, longitudinal distance, lateral distance and other information, the deep belief network DBN (Deep Belief A decision model based on a deep belief network (DBN) is used to control whether the vehicle should maintain centering or change lanes. The output of the decision model can be output to the planning module. Under the premise of vehicle lane change, a quintic polynomial is used to plan the lane change trajectory. At the same time, the three evaluation indicators of lane change efficiency, lane change comfort, and lane change safety are considered during the lane change process. A multi-objective lane change trajectory evaluation function is established, and the optimal lane change trajectory is solved based on a genetic algorithm. The planned lateral and longitudinal speeds and accelerations of the vehicle are then input into the control model, and the lateral and longitudinal control of the vehicle is achieved through model predictive control (MPC). During the simulation test, the autonomous driving system and the simulation software continuously exchange data, including information on the vehicle's motion state, the motion state of surrounding target vehicles, static road information, and other data. Data interaction is achieved through TCP (Transmission Control Protocol) communication.

[0055] According to the method for verifying the functional safety of autonomous driving simulation testing proposed in an embodiment of the present application, a scenario library for autonomous driving simulation testing is established, logical scenarios corresponding to the scenario library are automatically generated based on the complete parameters in a preset template, the design operation domain (ODD) of the autonomous driving simulation test is determined based on the logical scenarios based on a preset static map library, preset evaluation indicators, and a preset measurable scenario description language. Based on the ODD of the autonomous driving simulation test, the logical scenarios are generalized to obtain multiple target scenarios, and the algorithm to be verified of the vehicle under test is simulated and tested in each target scenario to obtain the safety results of the autonomous driving simulation test function. This method solves the problems of long time cycles, high investment costs, and low development efficiency in real-vehicle testing of autonomous driving systems, the inefficiency of scenario and simulation scenario generation methods in related technologies, the inability to automatically generate effective test scenarios in terms of scenario generalization, and the inability to automatically analyze the monitored vehicle behavior parameter data. It can realize the vision of high efficiency, high quality, and low cost of autonomous driving simulation testing, and has the advantages of scenario standardization, automated testing, and big data analysis.

[0056] Next, a device for verifying the functional safety of an autonomous driving simulation test proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0057] Figure 5 It is a block diagram of a device for verifying the safety of an autonomous driving simulation test function according to an embodiment of the present application.

[0058] like Figure 5 As shown, the verification device 10 for the functional safety of the autonomous driving simulation test includes: an establishment module 100, a determination module 200 and a testing module 300.

[0059] Among them, the establishment module 100 is used to establish a scenario library for autonomous driving simulation testing, and automatically generate logical scenarios corresponding to the scenario library based on the complete parameters in the preset template; the determination module is used to determine the design operation domain ODD of the autonomous driving simulation test according to the logical scenario based on the preset static map library, preset evaluation indicators and preset measurable scenario description language; the testing module is used to generalize the logical scenario based on the ODD of the autonomous driving simulation test to obtain multiple target scenarios, and simulate and test the algorithm to be verified of the vehicle to be tested in each target scenario to obtain the safety results of the autonomous driving simulation test function.

[0060] Furthermore, in some embodiments, after obtaining the safety results of the autonomous driving simulation test function, it also includes: an extraction module for extracting the target scenario where the simulation test fails from the safety results; an acquisition module for obtaining the motion state information of the vehicle to be tested under the target scenario where the simulation test fails; and a generation module for generating the reason for the failure of the simulation test based on the motion state information of the vehicle to be tested under the target scenario where the simulation test fails.

[0061] Furthermore, in some embodiments, the test module 300 is specifically used to: obtain a random seed corresponding to the current target scene; and store the motion state information of the current vehicle in the current target scene generated by the random seed corresponding to the current target scene to the target location.

[0062] Furthermore, in some embodiments, a logical scene corresponding to the scene library is automatically generated based on the complete parameters in the preset template, including: obtaining the Ego vehicle speed, the non-player character NPC vehicle speed, the longitudinal distance between the Ego vehicle speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scene, road geometry information, and road facility information; and automatically generating a logical scene corresponding to the scene library based on at least one parameter space among the Ego vehicle speed, the NPC vehicle speed, the longitudinal distance between the Ego vehicle speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scene, the road geometry information, and the road facility information.

[0063] Furthermore, in some embodiments, the scenario library includes at least one of a highway scenario library, an accident scenario library, an expert scenario library, and an extreme weather scenario library.

[0064] It should be noted that the above explanation of the embodiment of the method for verifying the safety of the autonomous driving simulation test function is also applicable to the verification device for the safety of the autonomous driving simulation test function of this embodiment, and will not be repeated here.

[0065] The device for verifying the functional safety of autonomous driving simulation testing, as proposed in an embodiment of the present application, establishes a scenario library for autonomous driving simulation testing, automatically generates logical scenarios corresponding to the scenario library based on complete parameters in a preset template, determines the design and operation domain (ODD) of the autonomous driving simulation test based on the logical scenarios, and generalizes the logical scenarios based on the ODD to obtain multiple target scenarios. The device then simulates and tests the algorithm to be verified on the vehicle under test in each target scenario to obtain the safety results of the autonomous driving simulation test function. This device solves the problems of long time cycles, high investment costs, and low development efficiency in real-vehicle testing of autonomous driving systems, the inefficiency of scenario and simulation scenario generation methods in related technologies, the inability to automatically generate effective test scenarios in scenario generalization, and the inability to automatically analyze monitored vehicle behavior parameter data. It achieves the vision of high efficiency, high quality, and low cost in autonomous driving simulation testing, and offers advantages such as scenario standardization, automated testing, and big data analysis.

[0066] Figure 6 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present application. The vehicle may include:

[0067] A memory 601 , a processor 602 , and a computer program stored in the memory 601 and executable on the processor 602 .

[0068] When the processor 602 executes the program, the verification method for the safety of the autonomous driving simulation test function provided in the above embodiment is implemented.

[0069] Furthermore, the electronic device further includes:

[0070] The communication interface 603 is used for communication between the memory 601 and the processor 602 .

[0071] The memory 601 is used to store computer programs that can be run on the processor 602 .

[0072] The memory 601 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.

[0073] If the memory 601, the processor 602, and the communication interface 603 are implemented independently, the communication interface 603, the memory 601, and the processor 602 can be connected to each other via a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0074] Optionally, in a specific implementation, if the memory 601, the processor 602 and the communication interface 603 are integrated on a chip, the memory 601, the processor 602 and the communication interface 603 can communicate with each other through an internal interface.

[0075] The processor 602 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0076] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for verifying the functional safety of the autonomous driving simulation test.

[0077] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0078] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature specified as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of this application, "N" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0079] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0080] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiment, the N steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array, a field programmable gate array, etc.

[0081] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0082] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for verifying functional safety of autonomous driving simulation testing, characterized in that: The following steps are involved: Establishing a scenario library for autonomous driving simulation testing, and automatically generating logical scenarios corresponding to the scenario library based on complete parameters in a preset template, wherein the scenario library includes at least one of a highway scenario library, an accident scenario library, an expert scenario library, and an extreme weather scenario library; The automatic generation of a logical scene corresponding to the scene library based on the complete parameters in the preset template includes: Obtaining the Ego's vehicle speed, the NPC's vehicle speed, the longitudinal distance between the Ego's vehicle speed and the NPC's vehicle speed, the NPC's vehicle lane, the duration of each stage in the logical scenario, road geometry information, and road facility information; Automatically generate a logical scenario corresponding to the scenario library based on at least one parameter space selected from the Ego vehicle speed, the NPC vehicle speed, the longitudinal distance between the Ego vehicle speed and the NPC vehicle speed, the NPC vehicle lane, the duration of each stage in the logical scenario, road geometry information, and road facility information; Based on a preset static map library, preset evaluation indicators, and a preset measurable scenario description language, a design operation domain (ODD) for the autonomous driving simulation test is determined according to the logical scenario. The preset evaluation indicators include statistical evaluation indicators, scenario-specific evaluation indicators, and multi-dimensional general evaluation indicators. The measurable scenario description language is based on Python, and the scenario trigger code "extend top.main" is used as the entry point for the automation program. The corresponding static map is selected through the "import . / hooder.xodr" method. Based on the ODD of the autonomous driving simulation test, the application security testing software Foretify is used to generalize the logical scenario to obtain multiple target scenarios. The principle of scenario generalization is based on relevant statistical theories and takes into account the vehicle dynamics model. Each logical scenario is generalized to generate at least one thousand different specific scenarios. Each specific scenario corresponds to a different random seed. The correspondence between the random seeds and specific scenarios is robust. The algorithm to be verified of the vehicle under test is simulated and tested in each target scenario to obtain the safety results of the autonomous driving simulation test function. After obtaining the safety result of the autonomous driving simulation test function, the method further includes: extracting target scenarios where simulation test failures occur from the safety results; Obtaining motion state information of the vehicle under test in a target scenario where the simulation test fails; The reason for the failure of the simulation test is generated based on the motion state information of the vehicle to be tested in the target scenario where the simulation test fails, including: finding the corresponding specific scenario through the corresponding random seed, and then analyzing the vehicle motion state information in the specific scenario to find the real cause of the failure, thereby further improving the automatic driving system of the vehicle to be tested.

2. The method according to claim 1, characterized in that After simulating and testing the algorithm to be verified of the vehicle to be tested in each target scenario, the method further includes: Get the random seed corresponding to the current target scene; The motion state information of the current vehicle in the current target scene generated by the random seed corresponding to the current target scene is stored in the target position.

3. A device for verifying functional safety of an autonomous driving simulation test for executing the method according to any one of claims 1 to 2, characterized in that: include: An establishment module is used to establish a scenario library for autonomous driving simulation testing and automatically generate logical scenarios corresponding to the scenario library based on complete parameters in a preset template, wherein the scenario library includes at least one of a highway scenario library, an accident scenario library, an expert scenario library, and an extreme weather scenario library; a determination module for determining a design operation domain (ODD) for an autonomous driving simulation test according to the logical scenario based on a preset static map library, preset evaluation indicators, and a preset measurable scenario description language, wherein the preset evaluation indicators include statistical evaluation indicators, scenario-specific evaluation indicators, and multi-dimensional general evaluation indicators; A testing module, configured to generalize the logic scenario to obtain multiple target scenarios based on the ODD of the autonomous driving simulation test, and simulate and test the algorithm to be verified of the vehicle under test in each target scenario to obtain a safety result of the autonomous driving simulation test function; After obtaining the safety result of the autonomous driving simulation test function, the method further includes: An extraction module, configured to extract target scenarios where simulation tests fail from the safety results; An acquisition module, configured to acquire motion state information of the vehicle under test in a target scenario where the simulation test fails; A generating module is used to generate a reason for the failure of the simulation test according to the motion state information of the vehicle to be tested in the target scenario where the simulation test fails.

4. The device according to claim 3, characterized in that The test module is specifically used to: Get the random seed corresponding to the current target scene; The motion state information of the current vehicle in the current target scene generated by the random seed corresponding to the current target scene is stored in the target position.

5. An electronic device, characterized in that: Including memory and processor; In which, the processor runs the program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the verification method for the functional safety of the autonomous driving simulation test as described in any one of claims 1-2.

6. A computer-readable storage medium storing a computer program, characterized in that: When the program is executed by a processor, it implements the method for verifying the functional safety of an autonomous driving simulation test as described in any one of claims 1-2.

Images