Method, apparatus, electronic device and storage medium for generating autonomous driving test scenarios

Abstract

The present invention relates to an automatic driving test scenario generation method, device, electronic device, and storage medium. The method includes: defining an automatic driving scenario according to a test standard; wherein, the defining of the automatic driving scenario includes: defining a logical scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form corresponding to the automatic driving scenario; setting parameters for the automatic driving scenario based on the forms; and generating an automatic driving test scenario based on the forms and the set parameters. The present invention defines scenarios according to test standards, which avoids the generation of scenarios that do not meet test requirements from the root cause; in addition, separating the work of scenario definition and parameter setting not only avoids a large number of similarities in scenario definition, but also facilitates the expansion and combination of parameters, reducing the number of scenarios that need to be independently defined and the number of duplicate scenarios.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving, and particularly to a method, device, electronic device, and storage medium for generating autonomous driving test scenarios. Background Art

[0002] Autonomous driving test scenarios are a very important part of an autonomous driving test system. The diversity, coverage, typicality, etc. of test scenarios can affect the accuracy of test results, thereby ensuring the safety and quality of autonomous driving.

[0003] Currently, when constructing autonomous driving test scenarios, traditional methods still use the scenario definition method of advanced driver assistance systems (ADAS) to define autonomous driving test scenarios.

[0004] However, the autonomous driving test scenarios constructed in this way have the following drawbacks:

[0005] (1) The traditional method of scenario definition focuses on changes in driving behavior and does not fully consider other influencing factors, which may result in test scenarios not meeting test requirements;

[0006] (2) Tens of thousands of test scenarios are required for autonomous driving tests. The scenario definition of traditional methods is based on test cases, and each test case describes 1 test scenario, and the number that needs to be independently defined is too large;

[0007] (3) In the scenario definition of traditional methods, the description main bodies of test cases for a type of test scenario are the same, and only the parameter parts are replaced, which is too repetitive;

[0008] (4) In the scenario definition of traditional methods, real vehicles and simulations are completely independent and do not have interoperability. Summary of the Invention

[0009] In view of the problems existing in the prior art, embodiments of the present invention provide an autonomous driving test scenario generation method, device, electronic device, and storage medium that overcome the above problems or at least partially solve the above problems.

[0010] In a first aspect, embodiments of the present invention provide an autonomous driving test scenario generation method, including:

[0011] Defining a function definition form, a logical scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to test standards;

[0012] Based on the forms, setting real vehicle / simulation test parameters;

[0013] Based on the forms and the set real vehicle / simulation test parameters, generating an autonomous driving scenario for real vehicle / simulation tests.

[0014] According to the method for generating an autonomous driving test scenario provided by the present invention, defining a function definition form, a logical scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to the test standard includes:

[0015] Interpret the test standard to determine the general attribute set of the autonomous driving scenario; wherein, the general attribute set includes: passing / triggering conditions, operational design domain, and dynamic driving tasks covered by the test standard;

[0016] Define a function definition form, a logical scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form based on the pre-acquired system / function design requirements and the general attribute set.

[0017] According to the method for generating an autonomous driving test scenario provided by the present invention, the function definition form includes: the number of the system / function design requirements, content description, associated autonomous driving scenario number, and the scenario category to which the associated autonomous driving scenario belongs;

[0018] The logical scenario form includes: the autonomous driving scenario number and its associated system / function design requirement number, dynamic driving tasks, passing / triggering conditions, host vehicle actions, host vehicle initial state, target vehicle actions, target vehicle initial state, and corresponding parameter descriptions;

[0019] The real vehicle parameter form includes: the autonomous driving scenario number, execution conditions and road geometry parameter pairs bound under the condition of no target vehicle and various types of target vehicles, and corresponding parameter descriptions;

[0020] The simulation parameter form includes: the autonomous driving scenario number, replacement parameters, designed operation domain, host vehicle actions, host vehicle initial state, target vehicle actions, target vehicle initial state, and corresponding parameter descriptions;

[0021] The signal form includes: the autonomous driving scenario number, signal names in the fd, and parameter descriptions corresponding to the manufacturer signal names;

[0022] Among them, the designed operation domain includes: road geometry, light intensity, and weather;

[0023] The initial state includes: vehicle type, initial speed, lane where the vehicle is located, vehicle head orientation, and location;

[0024] The actions include: action type, target speed, acceleration change, lane change direction, and lane change duration;

[0025] The corresponding parameter description is in the format of parameter name and an empty parameter value.

[0026] According to the method for generating an autonomous driving test scenario provided by the present invention, the pairs of execution conditions and road geometric parameters in the real vehicle parameter form are obtained by pre-selecting the execution conditions based on the test ratios under each road geometric type and then combining them;

[0027] The pairs of execution conditions and road geometric parameters bound under the non-target vehicle and each type of target vehicle are bound according to the test ratios of the non-target vehicle and each type of target vehicle.

[0028] According to the method for generating an autonomous driving test scenario provided by the present invention, the setting of real vehicle / simulation test parameters based on the form includes:

[0029] Summarize the parameters in the logical scenario form and the real vehicle parameter form with the same autonomous driving scenario number to obtain a set of real vehicle test parameters corresponding to the autonomous driving scenario number;

[0030] Summarize the parameters in the logical scenario form, the simulation parameter form, and the signal form with the same autonomous driving scenario number to obtain a set of simulation test parameters corresponding to the autonomous driving scenario number;

[0031] According to the setting rules of discrete parameters and continuous parameters, set the parameters for the set of real vehicle / simulation test parameters corresponding to the autonomous driving scenario number;

[0032] Among them, the discrete parameter setting rule is: parameter name, unit, and the variable can take discrete values; the continuous parameter setting rule is: parameter name, unit, variable value range, and variable step size.

[0033] According to the method for generating an autonomous driving test scenario provided by the present invention, the generation of an autonomous driving scenario for real vehicle test based on the form and the set real vehicle test parameters includes:

[0034] Based on the set of real vehicle test parameters corresponding to each autonomous driving scenario number, perform parameter setting, consider the function definition form, and automatically arrange, combine, and select the specific parameter values of the definition information in the logical scenario form and the real vehicle parameter form with the same autonomous driving scenario number to generate an autonomous driving test scenario for real vehicle test;

[0035] According to the method for generating an autonomous driving test scenario provided by the present invention, the generation of an autonomous driving scenario for simulation test based on the form and the set simulation test parameters includes:

[0036] Based on the set of simulation test parameters corresponding to each autonomous driving scenario number, parameter setting is performed. Considering the function definition form, the definition information of the logic scenario form, simulation parameter form, and signal form with the same autonomous driving scenario number is automatically arranged, combined, and the specific parameter values are selected to generate an autonomous driving test scenario for simulation testing.

[0037] In a second aspect, the present invention also provides an autonomous driving test scenario generation device, including:

[0038] A definition module, configured to: define a function definition form, a logic scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to test criteria;

[0039] A setting module, configured to perform real vehicle / simulation test parameter setting based on the forms;

[0040] A generation module, configured to generate an autonomous driving scenario for real vehicle / simulation testing based on the forms and the set real vehicle / simulation test parameters.

[0041] In a third aspect, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the autonomous driving test scenario generation method described in the first aspect are implemented.

[0042] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the autonomous driving test scenario generation method described in the first aspect are implemented.

[0043] The present invention defines an autonomous driving scenario including a function definition form, a logic scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to test criteria; fundamentally avoids the generation of autonomous driving scenarios that do not meet test requirements; performs real vehicle / simulation test parameter setting according to the forms; separates the work of scenario definition and parameter setting, which not only avoids a large number of similarities in scenario definition but also facilitates the expansion and combination of parameters; generates an autonomous driving scenario for real vehicle / simulation testing based on scenario definition and parameter setting, which not only solves the defect of an overly large number of scenarios that need to be independently defined but also solves the defect of overly repetitive generated scenarios. At the same time, the autonomous driving scenarios for real vehicle testing and simulation testing are connected in series through the logic scenario form, which is beneficial to the overall planning of the scenario library. Description of the Drawings

[0044] To more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0045] Figure 1 is a flowchart of a method for generating an autonomous driving test scenario provided by the present invention;

[0046] Figure 2 is a flowchart for establishing a scenario library provided by the present invention;

[0047] Figure 3 are the requirements for describing an autonomous driving test scenario provided by the present invention;

[0048] Figure 4 is a schematic diagram of the cloud parsing logic provided by the present invention;

[0049] Figure 5 is a structural diagram of a device for generating an autonomous driving test scenario provided by the present invention;

[0050] Figure 6 is a schematic structural diagram of an electronic device for implementing the method for generating an autonomous driving test scenario provided by the present invention. Detailed implementation manners

[0051] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention with reference to the accompanying drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0052] For ease of understanding, the following will explain the terms that appear in the embodiments of the present application.

[0053] A scenario is a comprehensive reflection of the environment and driving behavior within a certain time and space range, and is a driving condition jointly formed by external states such as weather conditions, roads, traffic facilities, and traffic participants, as well as information such as the driving tasks and states of the host vehicle.

[0054] Advanced Driver Assistance System (ADAS): is an active safety technology that uses various sensors installed on the vehicle to collect environmental data inside and outside the vehicle in real time, performs technical processing such as identification, detection, and tracking of static and dynamic objects, so that the driver can detect potential dangers in the shortest time, thereby attracting attention and improving safety.

[0055] Operational Design Domain (ODD): Refers to the operating conditions set for the functions of an autonomous driving system, including but not limited to environmental, geographical, and time - period restrictions, traffic flow, and road characteristics, etc.

[0056] Dynamic Driving Task (DDT): Refers to all real - time operations and strategic functions required to operate a vehicle on a road traffic, excluding functions such as trip arrangement, destination, and waypoint selection.

[0057] System / Function Design Requirements (RM): Include: Design requirements for ensuring the lateral movement of the vehicle within its own lane, and avoiding collisions with the vehicle in front that cuts into the lane in a critical state, etc.

[0058] Adjustable Damping System (ADS): Can be adjusted by the driver according to personal preferences, road conditions, and usage conditions to adjust the softness and hardness of the shock absorbers to suit different needs. By changing the damping force of the shock absorbers, a harder mode has a greater damping force to enhance the damping force during intense driving, and a softer mode provides a lower damping force for a more comfortable ride. Advanced adjustable shock absorber systems use electronic stepless adjustable shock absorber systems, which can actively and automatically adjust the most suitable shock absorber damping force according to different road conditions and operating conditions. However, due to the high price of this set of systems, they are usually only equipped in high - end luxury cars. The adjustable damping system can not only improve comfort but also contribute to driving safety.

[0059] Functional scenarios are used for project definition, risk analysis, and risk assessment in the concept stage. They describe the entities within the scenario area and the relationships between entities through language - scenario symbols.

[0060] Logical scenarios: Used to generate requirements during the project development stage. They express the entity characteristics and the relationships between entities by defining the parameter ranges of variables within the state space.

[0061] Specific scenarios can be directly used as test cases to clearly describe the entities and the relationships between entities by determining the specific values of each parameter in the state space.

[0062] OpenX (formerly phpAdsNew) is an advertising management and tracking system developed in PHP, suitable for various websites. It can manage multiple banner ads of any size owned by each advertiser, view statistics by day, detailed and summary statistics, and send reports to advertisers via email.

[0063] The following combines Figures 1-6 Describe the method, device, electronic device, and storage medium for generating an autonomous driving test scenario provided by the present invention.

[0064] In a first aspect, the present invention provides a method for generating an autonomous driving test scenario, as Figure 1 shown, the method comprising:

[0065] S11. Define a function definition form, a logical scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to the test standard;

[0066] Considering that traditional test case definitions come from behavioral patterns and do not comply with the scenario description requirements defined by the test standard, the present invention is committed to generating autonomous driving scenarios for simulation testing and real vehicle testing respectively based on the input of the test standard, avoiding the generation of autonomous driving scenarios that do not meet the test requirements from the source;

[0067] Different from the way of defining test cases (autonomous driving scenarios) with a single form in ADAS, the present invention uses a multi-form combination method to define test cases, which consists of five forms: a logical scenario form (case_lib), a function definition form (fd_lib), a real vehicle parameter form (para_veh_lib), a simulation parameter form (para_sim_lib), and a signal form (signal_lib);

[0068] Among them, the logical scenario form (case_lib) is the main body of the scenario definition part, which clarifies the test scenarios to be covered by each function, test execution, passing conditions, and the corresponding relationship with the function.

[0069] The function definition form (fd_lib) is a system definition of the function and one of the bases for scenario design. It includes a function definition number, a function definition description, and an associated scenario number. The purpose is to sort out the function requirements and achieve full coverage of the function requirements by the test scenarios.

[0070] The real vehicle parameter form (para_veh_lib) is used to define the parameters required for real vehicle testing and is filled in according to preset rules to facilitate subsequent correct permutation and combination of the parameters.

[0071] The simulation parameter form (para_sim_lib) is used to define the parameters required for simulation. In order to achieve subsequent automatic import of parameters and generation of simulation scenarios, the scenario needs to be decomposed into several specific variables, and the parameters are concatenated by establishing certain rules.

[0072] The signal form (signal_lib) is used to define the test scenarios of signal types, list the required signal names and signal values in different scenarios, and then, combined with the action sequence in the simulation parameter form, realize the sequential triggering of signals.

[0073] Through the multi-form combination definition method, a large number of similarities in scenario definition are avoided, and the number of scenarios that need to be independently defined is sharply reduced

[0074] S12. Set the real vehicle / simulation test parameters based on the form.

[0075] In the present invention, the scenario definition (i.e., the combined definition of five forms) and parameter setting are separated, which is beneficial to the accuracy of the scenario itself in the scenario definition stage and also facilitates the implementation of multiple specific test cases through the permutation and combination of different parameters. It can be understood that the parameter setting is to specify discrete optional values or continuous value ranges for the parameters in the scenario definition stage.

[0076] The work of scenario definition and parameter setting can both be completed in Excel. Of course, it can also be completed in the form of establishing documents, databases, etc., which is not limited here.

[0077] S13. Generate an autonomous driving scenario for real vehicle / simulation test based on the form and the set real vehicle / simulation test parameters.

[0078] The present invention arranges, combines and replaces numerical values for the defined information in the form by means of automation, and finally generates specific simulation test cases; by formulating scenario establishment rules mainly covering three stages of scenario definition, scenario parameter setting and scenario generation, the defects existing in the prior art are solved.

[0079] The present invention defines an autonomous driving scenario including a function definition form, a logical scenario form, a real vehicle parameter form, a simulation parameter form and a signal form according to the test standard; avoids the generation of autonomous driving scenarios that do not meet the test requirements from the root cause; sets the real vehicle / simulation test parameters according to the form; separates the work of scenario definition and parameter setting, which not only avoids a large number of similarities in scenario definition but also facilitates the expansion and combination of parameters; generates an autonomous driving scenario for real vehicle / simulation test on the basis of scenario definition and parameter setting, which not only solves the defect of too large a number of scenarios that need to be defined independently but also solves the defect of too repetitive generated scenarios. At the same time, the logical scenario form connects the autonomous driving scenarios for real vehicle test and the autonomous driving scenarios for simulation test, which is beneficial to the overall planning of the scenario library.

[0080] On the basis of the above embodiments, as an optional embodiment, the defining of the function definition form, the logical scenario form, the real vehicle parameter form, the simulation parameter form and the signal form according to the test standard includes:

[0081] Interpret the test standard to determine the general attribute set of the autonomous driving scenario; wherein, the general attribute set includes: pass / trigger conditions, operating design domain and dynamic driving tasks covered by the test standard.

[0082] It can be understood that the object of the present invention is to batch generate autonomous driving scenarios for simulation testing and real vehicle testing respectively with the test standard as the input. However, general test standards do not clearly guide how to establish scenarios. Considering that the key factors for building autonomous driving scenarios are passing / triggering conditions, operational design domain, and dynamic driving tasks, therefore, the test standard is interpreted, the key factors covered by the test standard are read, and then the scenarios are defined based on the key factors, so as to eliminate the generation of autonomous driving scenarios that do not meet the test requirements at the source.

[0083] Based on the pre-obtained system / function design requirements and general attribute set, define the function definition form, logical scenario form, real vehicle parameter form, simulation parameter form, and signal form.

[0084] It should be understood that when the present invention defines scenarios, it also needs to consider the development requirements of the autonomous driving test system (i.e., functional requirements or system / function design requirements) to ensure that the generated scenarios are not divorced from reality.

[0085] It should be clear that when actually building the scenario library, the construction of the scenario library must be applicable to different autonomous driving functional requirements. The construction plan of the scenario library can be summarized as "four stages" and "five in one". Among them, the four stages mainly include: the test standard interpretation stage, the scenario definition stage, the parameter setting stage, and the case generation stage. The five in one mainly includes: the logical scenario form (case_lib), the function definition form (fd_lib), the real vehicle parameter form (para_veh_lib), the simulation parameter form (para_sim_lib), and the signal form (signal_lib); Figure 2 The flowchart of the scenario library establishment is exemplified. This figure shows the last three stages from the second to the fourth of the scenario library construction, with different focuses in different stages. Separating the definition of the scenario from the setting of parameters is beneficial to the accuracy of the scenario itself in the scenario definition stage and also facilitates the realization of multiple specific cases by arranging and combining different parameters. Among them, the definition work in the second and third stages is completed in excel, following the pre-unified scenario library establishment rules. In the fourth stage, through automated means, the defined information in the table is arranged, combined, and replaced, and finally specific simulation cases are generated.

[0086] Based on the above embodiments, as an optional embodiment, the function definition form includes: the number of the system / function design requirements, the content description, the associated autonomous driving scenario number, and the scenario category to which the associated autonomous driving scenario belongs;

[0087] The described logical scenario form includes: the autonomous driving scenario number and its associated system / function design requirement number, dynamic driving tasks, passing / triggering conditions, actions of the host vehicle, initial states of the host vehicle, actions of the target vehicle, initial states of the target vehicle, and corresponding parameter descriptions;

[0088] The described vehicle parameter form includes: the autonomous driving scenario number, execution conditions and pairs of road geometry parameters bound under the condition of no target vehicle and various types of target vehicles, and corresponding parameter descriptions;

[0089] The described simulation parameter form includes: the autonomous driving scenario number, replacement parameters, designed operating domain, actions of the host vehicle, initial states of the host vehicle, actions of the target vehicle, initial states of the target vehicle, and corresponding parameter descriptions;

[0090] The described signal form includes: the autonomous driving scenario number, signal names in the fd, and parameter descriptions corresponding to the manufacturer signal names;

[0091] Among them, the designed operating domain includes: road geometry, light intensity, and weather;

[0092] The described initial state includes: vehicle type, initial speed, lane where the vehicle is located, vehicle head orientation, and location;

[0093] The described actions include: action type, target speed, acceleration change, lane change direction, and lane change duration;

[0094] The so-called parameter description is in the format of parameter name and an empty parameter value.

[0095] It should be noted that this embodiment only gives the preferred definition methods of the logical scenario form (case_lib), function definition form (fd_lib), vehicle parameter form (para_veh_lib), simulation parameter form (para_sim_lib), and signal form (signal_lib). In actual applications, it can be adjusted according to the working conditions. By defining the above 5 forms, a large number of similarities in scenario definitions are avoided, and the number of scenarios that need to be defined independently is sharply reduced.

[0096] In the present invention, forms other than the function definition form (fd_lib) all define the autonomous driving scenario number. Parameter aggregation and permutation and combination of definition information are carried out among the same autonomous driving scenario numbers, laying a foundation for subsequent parameter setting and scenario generation. In addition, by logically analyzing the logical scenario form (case_lib), function definition form (fd_lib), vehicle parameter form (para_veh_lib), simulation parameter form (para_sim_lib), and signal form (signal_lib), the permutation and combination of the definition information in the forms can be facilitated.

[0097] Based on the above embodiments, as an alternative embodiment, the pairs of execution conditions and road geometry parameters in the real vehicle parameter form are obtained by pre-selecting execution conditions based on the test ratios under each road geometry type and then combining them;

[0098] The pairs of execution conditions and road geometry parameters bound under the non-target vehicle and each type of target vehicle are bound according to the test ratios of the non-target vehicle and each type of target vehicle.

[0099] The execution condition (excution) variable and the road geometry (roadGeo) variable of the present invention form "parameter pairs" in pairs; several "parameter pairs" are bound under each scenario, so that the test ratios of the real vehicle test for the same scenario under different road geometries (roadGeo) vary according to requirements. For example, this scenario has been carefully tested under the road geometry (roadGeo) of a straight road, and only some targeted tests will be carried out on curves or slopes; in addition to the parameter pairs in the road geometry (roadGeo) dimension, by distinguishing different operating design domains (odd), the test ratios of the same scenario under different operating design domains (odd) also vary according to requirements. For example, this scenario has been carefully tested under the operating design domain (odd) of a sedan target, and only some targeted tests will be carried out under the operating design domain (odd) of a truck target;

[0100] Through such settings, the present invention ensures the practicability and flexibility of the autonomous driving test system.

[0101] Based on the above embodiments, as an alternative embodiment, the setting of real vehicle / simulation test parameters based on the form includes:

[0102] Summarize the parameters in the logical scenario form and the real vehicle parameter form with the same autonomous driving scenario number to obtain a set of real vehicle test parameters corresponding to the autonomous driving scenario number;

[0103] Summarize the parameters in the logical scenario form, the simulation parameter form and the signal form with the same autonomous driving scenario number to obtain a set of simulation test parameters corresponding to the autonomous driving scenario number;

[0104] Set the parameters of the real vehicle / simulation test parameter set corresponding to the autonomous driving scenario number according to the discrete parameter and continuous parameter setting rules;

[0105] Among them, the discrete parameter setting rule is: parameter name, unit, and the variable can take discrete values; the continuous parameter setting rule is: parameter name, unit, variable value range, and variable step size.

[0106] It can be understood that parameter setting is to define discrete optional values or continuous value ranges for the parameters specified in the scenario definition phase. For example, for the parameter setting of continuous parameters: speed; kph; 40 - 80; 10; for the parameter setting of discrete parameters: speed; kph; 40&50&80; speed represents the initial speed.

[0107] The present invention emphasizes "practicality". Different from the astronomical number of hundreds of millions of scenario libraries, for real vehicle tests, typical scenarios under typical parameters are adopted, and for simulation tests, a certain degree of traversal testing is adopted; that is, when setting parameters for real vehicle tests, they should be generated from real-world data, and when setting parameters for simulation tests, it is necessary to ensure that the selected value range is normalized;

[0108] In addition, the parameter sets for real vehicle / simulation tests contain multiple subsets, each subset corresponding to an autonomous driving scenario number. The autonomous driving scenarios for real vehicle tests and those for simulation tests are connected in series through the logical scenario forms with the same number, which is conducive to the overall planning of the scenario library.

[0109] Based on the above embodiments, as an optional embodiment, generating an autonomous driving scenario for real vehicle test based on the form and the set real vehicle test parameters includes:

[0110] Based on the real vehicle test parameter sets corresponding to each autonomous driving scenario number, perform parameter setting, consider the function definition form, and automatically arrange, combine, and select the specific parameter values of the definition information of the logical scenario form and the real vehicle parameter form with the same autonomous driving scenario number to generate an autonomous driving test scenario for real vehicle test;

[0111] The present invention can design a simple UI for different case designers to automatically analyze the function definition form, logical scenario form, and real vehicle parameter form with the same or related autonomous driving scenario numbers according to a certain logic, and then realize the arrangement and combination of the definition information in the logical scenario form and the real vehicle parameter form corresponding to the autonomous driving scenario number, laying a foundation for the generation of specific test cases.

[0112] In addition, the present invention complies with industry specifications, adopts a "three-layer model" of logical scenarios, parameter scenarios, and instantiated scenarios, and can output real vehicle test scenarios in the open X format.

[0113] Based on the above embodiments, as an optional embodiment, generating an autonomous driving scenario for simulation test based on the form and the set simulation test parameters includes:

[0114] Based on the set of simulation test parameters corresponding to each autonomous driving scenario number, parameter setting is performed. Considering the function definition form, the definition information of the logic scenario form, simulation parameter form, and signal form with the same autonomous driving scenario number is automatically arranged, combined, and the specific parameter values are selected to generate an autonomous driving test scenario for simulation testing.

[0115] The present invention can design a simple UI for different case designers to automatically parse the function definition form, logic scenario form, simulation parameter form, and signal form with the same or related autonomous driving scenario numbers according to a certain logic, thereby realizing the arrangement and combination of the definition information in the logic scenario form, simulation parameter form, and signal form corresponding to the autonomous driving scenario number, laying a foundation for the generation of specific test cases.

[0116] In addition, the present invention complies with industry specifications, adopts a "three-layer model" of logical scenarios, parameter scenarios, and instantiated scenarios, and outputs real vehicle test scenarios in the open X format.

[0117] To better explain the present invention, taking the ISO / CD 34501 Road vehicles — Terms and definitions of test scenarios for automated driving systems standard as an example, the generation process of the autonomous driving test scenario of the present invention is described as follows:

[0118] The present invention can be divided into four stages. In the first stage, it is necessary to interpret the input test standard, clarify the general attributes of the scenario, and understand the requirements of the scenario description. In the second and third stages, definitions are filled in Excel. Among them, in the second stage, the logic scenario form (case_lib), function definition form (fd_lib), real vehicle parameter form (para_veh_lib), simulation parameter form (para_sim_lib), and signal form (signal_lib) are defined. In the third stage, the relevant parameters for real vehicle / simulation testing are set and saved in Excel form. In the fourth stage, a Python script is written to parse multiple forms in Excel according to a certain logic, thereby generating specific test cases. Finally, a simple page design (web UI) is built through the Django framework (a simple Python web framework) for different case designers to use.

[0119] The specific process is as follows:

[0120] The first step: The test standard interpretation stage

[0121] Input the ISO / CD 34501 Road vehicles — Terms and definitions of test scenarios for automated driving systems standard. Since the testing and verification of adjustable damping systems (ADS) are one of the main challenges for market introduction, a six-layer model is defined to construct the surrounding environment of vehicles equipped with adjustable damping systems, which enables the definition of scenarios and the variation of influencing factors. Therefore, a general set of attributes for automated driving scenarios that can be used as input information for the six-layer model must be interpreted from this standard; the attributes of the general set of attributes can be used for oral scenario descriptions or for the requirements of other derived scenario description formats, such as those of OpenX. In addition, a method for classifying scenarios is proposed in this standard, and "tags" are defined as meta-attributes that provide additional information sources for each scenario.

[0122] This standard requires meeting the requirements for the description of automated driving test scenarios such as Figure 3 the examples. Therefore, the general set of attributes should at least include: pass / trigger conditions covered by the test standard, operational design domain, and dynamic driving tasks.

[0123] Step 2: Scenario definition phase

[0124] An automated driving test system has specific development requirements, namely system / function design requirements. The present invention defines scenarios based on the general set of attributes and system / function design requirements; among them, the scenario definition specifically executes the definition of the logical scenario form (case_lib), function definition form (fd_lib), real vehicle parameter form (para_veh_lib), simulation parameter form (para_sim_lib), and signal form (signal_lib).

[0125] 1. Definition of the function definition form (fd_lib):

[0126] The function definition form is the function definition of the automated driving test system and one of the bases for scenario design. It includes the function definition number, function definition description, and associated scenario number. The purpose is to sort out the requirements and achieve full coverage of the requirements by the test scenarios.

[0127] Among them, the structure of the function definition form can be as follows:

[0128]

[0129] Among them, represents the system / function design requirements, and here mainly fill in the number, Indicates the specific content of the system / function design requirements corresponding to RM. Indicates the scenarios related to RM. Here, mainly fill in the numbers of the scenarios related to RM. Indicates the filtering conditions, that is, to determine whether the scenarios related to RM are passed / triggered condition type scenarios or other type scenarios.

[0130] 2. Definition of the logical scenario form (case_lib):

[0131] The logical scenario form is the main body of the scenario definition part, which clarifies the test scenarios, test executions, passing conditions, and corresponding relationships with functional requirements that each function needs to cover. Among them, the test scenarios are divided into two expression methods: logical scenarios and parameter scenarios. The former focuses on the simple generalization of the scenarios, and the latter focuses on abstracting the parameters in the scenarios. The logical scenario form (case_lib) is the cornerstone of the parameter concretization in the second stage, and the scenario parameterization of the simulation will be instantiated based on the descriptions in the logical scenario form (case_lib).

[0132] Among them, the structure of the logical scenario form can be as follows:

[0133]

[0134] Among them, Indicates the number of the autonomous driving scenario. Indicates The associated system / function design requirements. Indicates the description of the dynamic driving task, such as speed limit, following, acceleration, etc., which is a specific manifestation of the test execution. Indicates the parameter description, which is the parameter of the scenario abstracted from this form. The parameter description includes: variable name and empty variable value. Indicates the initial state of the host vehicle. Indicates the action of the host vehicle. Indicates the initial state of the target vehicle. The action of the target vehicle. Of course, if there is more than one target vehicle, the initial states and actions of the target vehicles need to be listed one by one. Indicates the passing / triggering condition.

[0135] Definition of the real vehicle parameter form (para_veh_lib):

[0136] The real vehicle parameter form is used to define the parameters required for real vehicle testing and is filled in according to the following preset rules to facilitate the correct permutation and combination of the parameters later. When designing the form, in order to meet the timeliness and flexibility of real vehicle testing, the following rules are defined:

[0137] Rule 1: Distinguish between execution condition (excution) variables and road geometry (roadGeo) variables, and form "parameter pairs" pairwise;

[0138] Rule 2: Bind several "parameter pairs" under each scenario to achieve different test ratios for the same scenario under different road geometries (roadGeo) during real vehicle testing based on requirements. For example, if detailed tests are conducted on the road geometry (roadGeo) of a straight road for this scenario, only some targeted tests will be conducted on curves or slopes;

[0139] Rule 3: In addition to the parameter pairs in the road geometry (roadGeo) dimension, different operating design domains (odd) are distinguished through columns to achieve different test ratios for the same scenario under different operating design domains (odd) based on requirements. For example, if detailed tests are conducted on the operating design domain (odd) of a sedan for this scenario, only some targeted tests will be conducted on the operating design domain (odd) of a truck;

[0140] Rule 4: The variable format is "variable name: parameter value 1; parameter value 2; parameter value 3;..."; these parameter values are empty;

[0141] In view of this, the structure of the real vehicle parameter form can be as follows:

[0142]

[0143] Among them, represents the autonomous driving scenario number, represents the parameter description, which is the parameter of the scenario abstracted from this form. The parameter description includes: variable name and empty variable values, represents no target vehicle, represents the execution condition variable in the parameter pair bound to the no target vehicle, represents the road geometry variable in the parameter pair bound to the no target vehicle, represents the sedan target vehicle, represents the truck target vehicle. The sedan target vehicle and the truck target vehicle also have bound execution condition variable - road geometry variable parameter pairs, which are not shown in the simplified structure; the target vehicles under the scenario actually need to be listed one by one.

[0144] Definition of the simulation parameter form (para_sim_lib):

[0145] The simulation parameter form is used to define the parameters required for simulation. In order to achieve subsequent automatic parameter import and generation of simulation scenarios, the scenario needs to be decomposed into several specific variables, and parameter concatenation is achieved by establishing certain rules to generate the scenario.

[0146] The requirements of simulation tests are different from those of real vehicle tests. High flexibility is not required, but comprehensive coverage is demanded. Therefore, different from the design of "parameter pairs" in the real vehicle parameter form, the rules of the simulation parameter form are as follows:

[0147] Rule I: Starting from the simulation tool (VTD), sort out a total of 208 sets of simulation parameters to meet the definition requirements of different functions and different scenarios;

[0148] Rule II: According to the simulation tool's process of building a virtual scene, for the three objects of the host vehicle hv, the first target vehicle tv1, and the second target vehicle tv2, define the initial state (Init), action 1 (action1), action 2 (action2), action 3 (action3), and action 4 (action4) respectively;

[0149] Rule III: In the initial state, focus on defining information such as vehicle type, initial speed, lane where the vehicle is located, vehicle head orientation, and position where the vehicle is located. In the action description, focus on defining action type, target speed, acceleration change, lane change direction, lane change duration, etc.

[0150] Rule IV: In order to reflect the relationship with the operational design domain (odd), in addition to the simulation parameter information, additional variables such as road geometry (roadGeo) variables, illumination intensity (illumination) variables, and weather (weather) variables are added.

[0151] Rule V: The variable structure is "variable; unit; variable value or range; variable step size". Different from the variable structure in the real vehicle parameter form, the expanded step size information is introduced, which can be used for the traversal expansion of variables

[0152] In view of this, the structure of the simulation parameter form can be as follows:

[0153]

[0154] Among them, represents the autonomous driving scenario number, represents the parameter description, which is the parameter of the scenario abstracted from this form. The parameter description includes: variable name and empty variable range, represents the replacement parameter, represents the design operation domain, represents the host vehicle parameters, represents the parameters of the first target vehicle, represents the parameters of the second target vehicle. The target vehicle / host vehicle parameters define the initial state (Init) and actions (action) respectively. When defining the design operation domain, initial state (Init), and actions (action), refer to Table 1:

[0155] Table 1

[0156]

[0157]

[0158] In the above form structure, only the road geometry (roadGeo) variable, illumination variable, and weather variable are listed in the design operation domain; only the type ( ), initial speed ( ), initial lane ( ), and direction ( ) are listed in the initial state; only the type ( ), initial speed ( ), arrival position ( ), and action duration ( ) are listed in the action. Although the simulation parameter form only lists the parameters related to the second target vehicle, the host vehicle and the first target vehicle need to be processed in the same way. In addition, an action sequence needs to be set for a series of actions of the host vehicle, the first target vehicle, and the second target vehicle to facilitate triggering during simulation.

[0159] 5. Definition of signal form (signal_lib):

[0160] The signal form is used to define the test scenarios of signal types, list the required signal names and signal values in different scenarios, and then combine the action sequence in the simulation parameter form to achieve the sequential triggering of signals.

[0161] Among them, the structure of the signal form can be as follows:

[0162]

[0163] Among them, represents the autonomous driving scenario number, represents the signal name in represents the manufacturer signal name, and both represent signal values.

[0164] The third step: Parameter setting stage

[0165] In the parameter setting stage, the parameters for real vehicle and simulation tests need to be summarized separately. That is, based on the same autonomous driving scenario number, the parameters in the logical scenario form and the real vehicle parameter form are summarized to obtain the parameters for real vehicle tests under each autonomous driving scenario number; based on the same autonomous driving scenario number, the parameters in the logical scenario form, the simulation parameter form, and the signal form are summarized to obtain the parameters for simulation tests under each autonomous driving scenario number;

[0166] Then, according to the discrete and continuous parameter setting rules shown in Table 2, parameter definitions are made for the parameters used in real vehicle / simulation tests under each autonomous driving scenario number;

[0167] Table 2

[0168]

[0169] This stage is mainly to assign value ranges to the parameters. In the present invention, the scenario definition and parameter setting are separated, which is beneficial to the accuracy of the scenario itself in the scenario definition stage and also facilitates the implementation of expanding into multiple specific cases through the permutation and combination of different parameters.

[0170] Step 4: Case generation stage

[0171] The case generation solution is implemented based on the Django web framework. The advantage is that developers deploy the algorithm in the cloud and the parsing work is completed in the cloud. Users do not need to pre-install the environment and can enable the parsing function in the cloud by directly accessing the web link. The implementation logic is as Figure 4 shown in the schematic diagram of the cloud parsing logic. This part supports single-function parsing and multi-function parsing.

[0172] Generate simulation test cases

[0173] After each functional user uploads the form to the cloud and clicks the "simulation csae" button, an intermediate file in the openX format will be generated, and this intermediate file can be used to convert into the xml file required for VTD simulation.

[0174] Generate scenario statistics

[0175] In addition to being able to statistically extract the number of logical scenarios in the scenario library, users can also statistically count the number of scenarios of different test methods, such as the number of scenarios that can be achieved through automated simulation means and the number of scenarios that require manual simulation.

[0176] In the second aspect, the autonomous driving scenario generation device provided by the present invention is described. The autonomous driving scenario generation device described below can be mutually referred to and corresponding to the autonomous driving scenario generation method described above. Figure 5 An example of the structural schematic diagram of an autonomous driving test scenario generation device is shown in Figure 5 as shown. This device includes: a definition module 21, a setting module 22, and a generation module 23;

[0177] Among them, a definition module 21 is used to: define a function definition form, a logic scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to test criteria; a setting module 22 is used to set real vehicle / simulation test parameters based on the forms; a generation module 23 is used to generate an autonomous driving scenario for real vehicle / simulation test based on the forms and the set real vehicle / simulation test parameters.

[0178] The autonomous driving test scenario generation device provided by the embodiment of the present invention specifically executes the content of each of the above-mentioned embodiments of the autonomous driving test scenario generation method, which will not be elaborated here. The autonomous driving test scenario generation device provided by the embodiment of the present invention defines an autonomous driving scenario including a function definition form, a logic scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to test criteria; fundamentally avoids the generation of autonomous driving scenarios that do not meet test requirements; sets real vehicle / simulation test parameters according to the forms; separates the work of scenario definition and parameter setting, which not only avoids a large number of similarities in scenario definition but also facilitates the expansion and combination of parameters; generates an autonomous driving scenario for real vehicle / simulation test based on scenario definition and parameter setting, which not only solves the defect of an overly large number of scenarios that need to be defined independently but also solves the defect of overly repetitive generated scenarios. At the same time, the logic scenario form connects the autonomous driving scenarios for real vehicle test and the autonomous driving scenarios for simulation test, which is beneficial to the overall planning of the scenario library.

[0179] In a third aspect, Figure 6 An entity structure diagram of an electronic device is exemplified, as Figure 6 shown. The electronic device may include: a processor 610, a communication interface 620, a memory 630, and a communication bus 640. Among them, the processor 610, the communication interface 620, and the memory 630 communicate with each other through the communication bus 640. The processor 610 can call the logical instructions in the memory 630 to execute the autonomous driving test scenario generation method, which includes: defining a function definition form, a logic scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to test criteria; setting real vehicle / simulation test parameters based on the forms; generating an autonomous driving scenario for real vehicle / simulation test based on the forms and the set real vehicle / simulation test parameters.

[0180] In addition, when the logical instructions in the above-mentioned memory 630 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0181] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is used to execute an autonomous driving test scenario generation method, and the method includes: defining a function definition form, a logical scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to test criteria; setting real vehicle / simulation test parameters based on the forms; and generating an autonomous driving scenario for real vehicle / simulation tests based on the forms and the set real vehicle / simulation test parameters.

[0182] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative labor.

[0183] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the technical solution, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disks, optical discs, etc., and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments.

[0184] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for generating an autonomous driving test scenario, characterized in that The method includes: Defining a function definition form, a logical scenario form, a real vehicle parameter form, a simulation parameter form, and a signal form according to test criteria; Based on the forms, setting real vehicle / simulation test parameters; Based on the forms and the set real vehicle / simulation test parameters, generating an autonomous driving scenario for real vehicle / simulation testing; The defining of the function definition form, the logical scenario form, the real vehicle parameter form, the simulation parameter form, and the signal form according to test criteria includes: Interpreting the test criteria to determine a general attribute set for the autonomous driving scenario; wherein, the general attribute set includes: passing / triggering conditions, operational design domain, and dynamic driving tasks covered by the test criteria; Based on the pre-acquired system / function design requirements and the general attribute set, defining the function definition form, the logical scenario form, the real vehicle parameter form, the simulation parameter form, and the signal form; The setting of real vehicle / simulation test parameters based on the forms includes: Summarizing the parameters in the logical scenario form and the real vehicle parameter form with the same autonomous driving scenario number to obtain a set of real vehicle test parameters corresponding to the autonomous driving scenario number; Summarizing the parameters in the logical scenario form, the simulation parameter form, and the signal form with the same autonomous driving scenario number to obtain a set of simulation test parameters corresponding to the autonomous driving scenario number; According to the discrete parameter and continuous parameter setting rules, setting the real vehicle / simulation test parameter sets corresponding to the autonomous driving scenario number; Wherein, the discrete parameter setting rule is: parameter name, unit, and the variable can take discrete values; the continuous parameter setting rule is: parameter name, unit, variable value range, and variable step size.

2. The method for generating an autonomous driving test scenario according to claim 1, wherein The function definition form includes: the number of the system / function design requirements, content description, associated autonomous driving scenario number, and the scenario category to which the associated autonomous driving scenario belongs; The logical scenario form includes: the autonomous driving scenario number and its associated system / function design requirement number, dynamic driving task, passing / triggering condition, host vehicle action, host vehicle initial state, target vehicle action, target vehicle initial state, and corresponding parameter description; The real vehicle parameter form includes: the autonomous driving scenario number, execution conditions and road geometry parameter pairs bound under the condition of no target vehicle and various types of target vehicles, and corresponding parameter description; The simulation parameter form includes: the autonomous driving scenario number, replacement parameters, designed operation domain, host vehicle action, host vehicle initial state, target vehicle action, target vehicle initial state, and corresponding parameter description; The signal form includes: the autonomous driving scenario number, signal names in the fd, and parameter descriptions corresponding to the manufacturer signal names; Wherein, the designed operation domain includes: road geometry, light intensity, and weather; The initial state includes: vehicle type, initial speed, lane where the vehicle is located, vehicle head orientation, and location; The action includes: action type, target speed, acceleration change, lane change direction, and lane change duration; The parameter description is in the format of parameter name and an empty parameter value.

3. The method for generating an autonomous driving test scenario according to claim 2, wherein The execution conditions in the actual vehicle parameter form are paired with the road geometric parameters, which are obtained by pre-selecting the execution conditions based on the test ratios under each road geometric type and then combining them; The execution condition and road geometric parameter pairs bound under the non-target vehicle and each type of target vehicle are bound according to the test ratios of the non-target vehicle and each type of target vehicle.

4. The method for generating an autonomous driving test scenario according to claim 1, wherein Based on the form and the set actual vehicle test parameters, the generation of an autonomous driving scenario for actual vehicle testing includes: Based on the actual vehicle test parameter sets corresponding to each autonomous driving scenario number, parameter setting is performed. Considering the function definition form, the definition information of the logical scenario form and the actual vehicle parameter form with the same autonomous driving scenario number is automatically arranged, combined, and specific parameter values are selected to generate an autonomous driving test scenario for actual vehicle testing.

5. The method for generating an autonomous driving test scenario according to claim 1, wherein Based on the form and the set simulation test parameters, the generation of an autonomous driving scenario for simulation testing includes: Based on the simulation test parameter sets corresponding to each autonomous driving scenario number, parameter setting is performed. Considering the function definition form, the definition information of the logical scenario form, the simulation parameter form, and the signal form with the same autonomous driving scenario number is automatically arranged, combined, and specific parameter values are selected to generate an autonomous driving test scenario for simulation testing.

6. An automatic driving test scenario generation device, characterized in that The device includes: A definition module for: defining a function definition form, a logical scenario form, an actual vehicle parameter form, a simulation parameter form, and a signal form according to the test standard; A setting module for performing actual vehicle / simulation test parameter setting based on the form; A generation module for generating an autonomous driving scenario for actual vehicle / simulation testing based on the form and the set actual vehicle / simulation test parameters; The definition module includes: Interpret the test standard to determine the general attribute set of the autonomous driving scenario; wherein, the general attribute set includes: passing / triggering conditions, operational design domain, and dynamic driving tasks covered by the test standard; Define a function definition form, a logical scenario form, an actual vehicle parameter form, a simulation parameter form, and a signal form based on the pre-obtained system / function design requirements and the general attribute set; The setting module includes: Summarize the parameters in the logical scenario form and the actual vehicle parameter form with the same autonomous driving scenario number to obtain the actual vehicle test parameter set corresponding to the autonomous driving scenario number; Summarize the parameters in the logical scenario form, the simulation parameter form, and the signal form with the same autonomous driving scenario number to obtain the simulation test parameter set corresponding to the autonomous driving scenario number; According to the discrete parameter and continuous parameter setting rules, perform parameter setting on the actual vehicle / simulation test parameter set corresponding to the autonomous driving scenario number; Among them, the discrete parameter setting rule is: parameter name, unit, and the variable can take discrete values; the continuous parameter setting rule is: parameter name, unit, variable value range, and variable step size.

7. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein, When the processor executes the program, it implements the steps of the autonomous driving test scenario generation method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the autonomous driving test scenario generation method according to any one of claims 1 to 5.

Images