CN110795818A

CN110795818A - Method and device for determining virtual test scene, electronic equipment and storage medium

Info

Publication number: CN110795818A
Application number: CN201910867084.1A
Authority: CN
Inventors: 孙驰天
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2019-09-12
Filing date: 2019-09-12
Publication date: 2020-02-14
Anticipated expiration: 2039-09-12
Also published as: CN110795818B

Abstract

The application relates to the technical field of automatic driving, in particular to a method, a device, electronic equipment and a storage medium for determining a virtual test scene, which are used for improving the construction efficiency and the authenticity of the virtual test scene of an automatic driving vehicle, wherein the method comprises the following steps: obtaining input scene configuration parameters, and determining a test map according to map information in the scene configuration parameters; configuring the test map according to the configuration information of the automatic driving vehicles in the scene configuration parameters to generate a virtual test scene corresponding to the test map, wherein the virtual test scene comprises at least one automatic driving vehicle and at least one scene element; performing virtual test on the automatic driving vehicle according to the virtual test scene; and determining the virtual test scene with the abnormality in the virtual test process as the virtual test scene of the automatic driving vehicle. Because the simulation test scene can be automatically generated in large batch, the construction efficiency and the authenticity of the scene of automatic driving are improved.

Description

Method and device for determining virtual test scene, electronic equipment and storage medium

Technical Field

The application relates to the computer technology, in particular to the technical field of automatic driving, and provides a method and a device for determining a virtual test scene, electronic equipment and a storage medium.

Background

The automatic driving vehicle senses the surrounding environment of the vehicle through a vehicle-mounted sensing system, and controls the steering and the speed of the vehicle according to road information, other vehicle position information, obstacle information and the like obtained through sensing, so that the vehicle can safely and reliably run on the road.

In the related art, an automatic driving test scenario is constructed through simulation software to perform simulation test on an automatic driving vehicle, and a common method for constructing the automatic driving test scenario is as follows: based on the self-defined scene of the scene editor, traditional simulation software such as VTD (Virtual Test Drive), PreScan, etc. are all provided with the scene editor for self-defining editing of the scene.

However, although it is convenient to generate a few scenes because the scene editor completely relies on user input to generate a scene, and the user selects input such as pedestrians, obstacles, and vehicles as shown in fig. 1, the automatic driving at a Level of L3(Level 3) or higher requires a scene library in units of ten thousand, and thus manual editing of a scene by the scene editor becomes an unrealistic option. Meanwhile, the manually defined scene element behaviors are not real enough, and have negative effects on the training result of the method.

Disclosure of Invention

The embodiment of the application provides a method, a device, electronic equipment and a storage medium for determining a virtual test scene of an automatic driving vehicle, which are used for improving the construction efficiency of the virtual test scene of the automatic driving vehicle and improving the authenticity of the virtual test scene of the automatic driving vehicle.

The method for determining the virtual test scene of the automatic driving vehicle comprises the following steps:

obtaining input scene configuration parameters, and determining a test map for carrying out virtual test on the automatic driving vehicle according to map information in the scene configuration parameters;

configuring the test map according to the configuration information of the automatic driving vehicles in the scene configuration parameters to generate a virtual test scene corresponding to the test map, wherein the virtual test scene comprises at least one automatic driving vehicle and at least one scene element;

performing virtual testing on the automatic driving vehicle according to the virtual testing scene; and are

And determining the virtual test scene with the abnormality in the virtual test process as the virtual test scene of the automatic driving vehicle.

The device for determining the virtual test scene of the automatic driving vehicle provided by the embodiment of the application comprises the following steps:

the system comprises a parameter acquisition unit, a parameter analysis unit and a parameter analysis unit, wherein the parameter acquisition unit is used for acquiring input scene configuration parameters and determining a test map for carrying out virtual test on the automatic driving vehicle according to map information in the scene configuration parameters;

the scene generating unit is used for configuring the test map according to the configuration information of the automatic driving vehicles in the scene configuration parameters to generate a virtual test scene corresponding to the test map, wherein the virtual test scene comprises at least one automatic driving vehicle and at least one scene element;

the virtual test unit is used for carrying out virtual test on the automatic driving vehicle according to the virtual test scene; and determining the virtual test scene with the abnormality in the virtual test process as the virtual test scene of the automatic driving vehicle.

Optionally, the virtual test scenario with the exception includes:

when the automatic driving vehicle is subjected to virtual testing, at least one automatic driving vehicle does not pass through a virtual testing scene of at least one working condition.

Optionally, the parameter obtaining unit is further configured to:

and acquiring an input trigger interval through the configuration interface, and increasing a set number of traffic flow vehicles in the virtual test scene at intervals of the trigger interval, wherein the traffic flow vehicles run according to traffic rules.

Optionally, the parameter obtaining unit is further configured to:

and acquiring the input aggressive degree parameter through the configuration interface, and adjusting the running state of the traffic flow vehicle according to the aggressive degree parameter.

An electronic device provided by an embodiment of the present application includes a processor and a memory, where the memory stores program code, and when the program code is executed by the processor, the processor executes the steps of the method for determining a virtual test scenario of an autonomous vehicle.

Embodiments of the present application provide a non-transitory computer-readable storage medium including program code for causing an apparatus for determining virtual test scenarios for autonomous vehicles to perform the steps of one of the above-described methods for determining virtual test scenarios for autonomous vehicles, when the program product is run on the apparatus for determining virtual test scenarios for autonomous vehicles.

The beneficial effect of this application is as follows:

the embodiment of the application provides a method, a device, electronic equipment and a storage medium for determining a virtual test scene of an automatic driving vehicle, and the embodiment of the application is based on a cloud simulation architecture, can automatically load a map of a large area according to map information input by a user, generates the virtual test scene according to vehicle configuration information input by the user and the like, runs large area simulation according to the generated virtual test scene, can meet various different traffic scenes when the automatic driving vehicle freely runs in the virtual area, and can use the abnormal virtual test scene as the virtual test scene of the automatic driving vehicle when an abnormal scene which cannot be processed occurs, so that meaningful simulation test scenes such as high speed, urban areas and the like can be automatically generated in large batch, the problem of coverage of the automatic driving scene is solved, and waste of a large amount of manpower, material resources and time is avoided, the problem of long tail effect of scene coverage which is difficult to solve by manual editing is well solved, the establishment of an automatic driving scene library is accelerated, the construction efficiency and the authenticity of an automatic driving scene are improved, and then the landing of an automatic driving product is accelerated.

Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the application. The objectives and other advantages of the application may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:

fig. 1 is an alternative schematic diagram of a scene editor in an embodiment of the present application;

fig. 2A is a schematic view of an application scenario according to an embodiment of the present application;

FIG. 2B is a schematic diagram of an alternative configuration of a virtual simulation system to which embodiments of the present application are applied;

FIG. 3 is an alternative flow chart illustrating a method for determining a virtual test scenario for an autonomous vehicle in an embodiment of the present application;

FIG. 4 is an alternative schematic diagram of a user interaction interface of a virtual simulation system in an embodiment of the present application;

FIG. 5 is an alternative diagram of simulation results of a virtual simulation system in an embodiment of the present application;

FIG. 6A is an alternative diagram of an exception scenario in an embodiment of the present application;

FIG. 6B is an alternative diagram of an exception scenario in an embodiment of the present application;

FIG. 7 is a schematic diagram of an alternative operational flow of a virtual simulation system in an embodiment of the present application;

FIG. 8 is a schematic illustration of an alternative map cut in an embodiment of the present application;

FIG. 9A is a schematic diagram of an alternative virtual test scenario in an embodiment of the present application;

FIG. 9B is a schematic diagram of another alternative virtual test scenario in an embodiment of the present application;

FIG. 10A is a schematic diagram of an alternative adjusted virtual test scenario in an embodiment of the present application;

FIG. 10B is a schematic diagram of an alternative adjusted virtual test scenario in the embodiment of the present application;

FIG. 11 is a schematic timing flow diagram illustrating an implementation of an alternative method for determining a virtual test scenario for an autonomous vehicle according to an embodiment of the present application;

FIG. 12 is a schematic structural diagram illustrating an apparatus for determining a virtual test scenario of an autonomous vehicle according to an embodiment of the present disclosure;

fig. 13 is a schematic diagram of a hardware component of a computing device to which an embodiment of the present application is applied.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments, but not all embodiments, of the technical solutions of the present application. All other embodiments obtained by a person skilled in the art without any inventive step based on the embodiments described in the present application are within the scope of the protection of the present application.

Some concepts related to the embodiments of the present application are described below.

1. L3: in 2014, SAE International (society of automated engineers) developed a set of automotive grading standards for automatic driving, which were classified into 5 grades for describing automation from full manual driving to full automatic driving. L3 indicates that the automated system is capable of both performing certain driving tasks and in some cases monitoring the driving environment, but that the driver must be ready to regain driving control.

2. Monte carlo method: also called statistical simulation method, statistical test method. The method is a numerical simulation method using a probability phenomenon as a research object. The method is a calculation method for estimating an unknown characteristic amount by obtaining a statistical value by a sampling survey method. Monte Carlo is a famous gamble in Morna, which is named to indicate its randomly sampled nature. Therefore, the method is suitable for performing calculation simulation tests on the discrete system. In computational simulation, the stochastic nature of the system can be simulated by constructing a probabilistic model that approximates the performance of the system and performing stochastic tests on a digital computer.

3. Long tail effect: the English name Long Tail Effect. The "head" and "tail" are two statistical terms. The projection in the middle of the normal curve is called the "head"; the relatively flat parts on both sides are called tail. From the perspective of human demand, most of the demand will be concentrated on the head, and this part we can call popular, while the demand distributed on the tail is personalized, scattered and small demand. This part of the differentiated, small demand creates a long "tail" on the demand curve, and the so-called long tail effect is in its number, adding up all non-popular markets creates a larger market than the popular market.

4. And (3) regression testing: after the old code is modified, the virtual test is performed again to confirm that the modification does not introduce new errors or cause errors in other codes. The automatic regression testing can greatly reduce the cost of the system testing, maintenance upgrading and other stages. The regression test is used as a component of the software life cycle, and occupies a great workload proportion in the whole software test process, and multiple regression tests can be carried out at each stage of software development. In progressive and fast iterative development, the continuous release of new versions makes regression testing more frequent, whereas in extreme programming methods it is even more required to perform regression testing several times per day. Therefore, it is very meaningful to improve the efficiency and effectiveness of regression testing by selecting the right regression testing strategy.

5. Traffic flow: the vehicle is a traffic flow formed by continuous driving of the vehicle on a road. But also traffic and people flows of other vehicles in a broad sense. In a certain period of time, on a road section which is not influenced by transverse intersection, the traffic flow is in a continuous flow state; when meeting the control of the signal lamp at the intersection, the traffic light is in a discontinuous flow state.

6. Working conditions are as follows: the method refers to the synthesis of the environment condition of the running of the automobile and the condition of the engine of the automobile, and can be subdivided into the running conditions of starting, accelerating, constant speed, decelerating, turning, ascending and descending, stopping and the like. The method mainly comprises the following steps of: the system comprises the working conditions of gear shifting, speed changing, sliding (out-of-gear sliding, neutral sliding, accelerating sliding and parking sliding), braking (emergency braking, speed control braking and brake braking), accelerator speed control, steering, reversing and the like. According to the load condition, the method mainly comprises the following steps: no-load, full load (equal to rated load), overload (exceeding rated load) and other operation conditions.

7. VTD: the visual simulation tool is a complex traffic scene visual simulation tool developed by Germany VIRES company, and can be applied to real-time complex traffic scene generation (including sensor imaging of infrared and cameras) of automobile active safety and unmanned automobile semi-physical test and traffic visual display in automobile driving simulator development.

8. PreScan: the simulation platform is a simulation platform based on physics, and is used for developing an ADAS (Advanced Driving assistance System) based on sensor technologies such as radar, laser/laser radar, camera and GPS (Global Positioning System) in the automobile industry. PreScan can also be used to design and evaluate vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication applications as well as autonomous driving applications.

The preferred embodiments of the present application will be described below with reference to the accompanying drawings of the specification, it should be understood that the preferred embodiments described herein are merely for illustrating and explaining the present application, and are not intended to limit the present application, and that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.

Artificial Intelligence (AI) is a theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and expand human Intelligence, perceive the environment, acquire knowledge and use the knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technique of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a manner similar to human intelligence. Artificial intelligence is the research of the design principle and the realization method of various intelligent machines, so that the machines have the functions of perception, reasoning and decision making.

The artificial intelligence technology is a comprehensive subject and relates to the field of extensive technology, namely the technology of a hardware level and the technology of a software level. The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, a machine learning/deep learning technology, an automatic driving technology and the like.

The automatic driving technology generally comprises technologies such as high-precision maps, environment perception, behavior decision, path planning, motion control and the like, and has wide application prospects.

According to the level of the automatic driving ability, the automatic driving can be mainly classified into 5-level, 1-level is driving assistance, 2-level is partial automatic driving, 3-level is conditional automatic driving, 4-level is highly automatic driving, and 5-level is full automatic driving. The automatic driving system with intelligent levels of 1 and 2 mainly depends on human beings to monitor the driving environment, and the automatic driving system with 3 to 5 levels monitors the driving environment through a computer system. With the improvement of the automatic driving intelligence level, the intelligent system gradually reaches and is provided with the driving level of a human driver, and when the automatic driving intelligence level reaches 3-5 levels, the driving task is completed in a part of scenes instead of the human.

With the current continuous and deep research on the automatic driving function by domestic and foreign research institutions, the intelligent level of the automatic driving vehicle gradually realizes the promotion from auxiliary driving to unmanned driving. In order to ensure the operation safety and stability of intelligent systems with different levels, virtual test authentication needs to be performed on the basic safety performance of the automatic driving system, and the operation safety of the automatic driving vehicle needs to be verified according to a test result.

The scheme provided by the embodiment of the application relates to an artificial intelligent automatic driving technology, and is specifically explained by the following embodiment:

fig. 2A is a schematic view of an application scenario in the embodiment of the present application. In the embodiment of the present application, the virtual city system may be displayed on the client 220, the user 210 inputs scene configuration parameters through the configuration interface 221 of the client 220, where the scene configuration parameters include map information and configuration information of the autonomous vehicles, the cloud server 230 is a server corresponding to the client 220, the cloud server 230 obtains the scene configuration parameters input by the user 210 through the client 220, then loads a test map for performing a virtual test on the autonomous vehicles according to the map information, and configures the test map according to the configuration information of the autonomous vehicles to generate a virtual test scene corresponding to the test map, where the virtual test scene includes at least one autonomous vehicle and at least one scene element. The cloud server 230 displays the virtual test scenario on the client 220, as shown by a simulation interface 222 on the client 220, the virtual test scenario is an optional virtual test scenario, after the virtual test scenario is generated, the autonomous vehicle is virtually tested by operating a virtual city system, and when an abnormality occurs in the virtual test scenario is detected, the virtual test scenario can be stored in a scenario library, so that the scenario coverage of autonomous driving is improved, and the scenario coverage long-tail effect which is difficult to solve by manual editing is solved.

The embodiment of the application is based on a cloud simulation architecture to construct a scene, wherein a virtual city system can be displayed on a client, and the virtual city system comprises an evaluation module 10 and a video recording module 20, as shown in fig. 2B.

The evaluation module 10 evaluates the autopilot algorithm in real time according to the traffic laws, rules and comfort conditions, and detects whether the virtual test scene is abnormal. Video recording module 20 records all behaviors of all autonomous vehicles and surrounding vehicles (traffic flow vehicles and/or autonomous vehicles) in real time. When the evaluation module 10 identifies a scene that is not processed by the autopilot algorithm configured for the autopilot vehicle in the virtual test scene (i.e., a scene violating the traffic laws and regulations or not meeting the comfort level condition), for example, when overspeed occurs, evaluation of the traffic laws and regulations is violated, and when sudden braking occurs, the comfort level condition is not met, in this case, it is detected that the virtual test scene is abnormal, the video module 20 intercepts and stores the video data for a preset time period before the time point when the abnormal occurs, so that extraction of the abnormal scene is completed.

The scenes violating the traffic laws and regulations include, but are not limited to, scenes of speeding, pressing lines, collision, abnormal approach to other vehicles, running red light and the like.

The condition that the comfort level is not met refers to the condition that the comfort level of passengers taking the automatic driving vehicle cannot be guaranteed, and scenes that the comfort level condition is not met include but are not limited to scenes such as sudden braking or sudden acceleration.

In the embodiment of the application, the video recording module records all behaviors of all automatic driving vehicles and surrounding vehicles (traffic flow vehicles and/or automatic driving vehicles) in real time. When the evaluation module identifies a scene (referring to a scene violating the traffic laws and regulations or not meeting the comfort condition) which cannot be processed by the automatic driving algorithm configured by the automatic driving vehicle in the virtual test scene, for example, when overspeed occurs, the evaluation of the traffic laws and regulations is violated, and when sudden braking occurs, the comfort condition is not met, and under the condition, the virtual test scene is detected to be abnormal, the video module intercepts and stores the video data with the preset time length before the abnormal time point, so that the extraction of the abnormal scene is completed.

Referring to fig. 3, an implementation flow chart of a method for determining a virtual test scenario of an autonomous vehicle according to an embodiment of the present application is shown, and a specific implementation flow of the method is as follows:

s31: the method comprises the steps of obtaining input scene configuration parameters, and determining a test map for carrying out virtual test on the automatic driving vehicle according to map information in the scene configuration parameters.

In an alternative schematic diagram of the configuration interface in the embodiment of the present application, reference may be made to fig. 4, where the current configuration interface is a task management/virtual city task/creation task interface.

The task management interface (configuration interface) can manage all tasks which are required to be run; the scene library is used for operating a set scene library and giving a method evaluation report based on the operation result of the scene in the scene library; the virtual city task can construct a virtual environment at a city level, thousands of automatic driving automobiles and millions of traffic flow vehicles can be continuously operated in the virtual environment for 24 hours, a virtual test scene is constructed to virtually test the automatic driving vehicles, various test scenes which are difficult to be imagined by human brains are triggered by the large random system, and the scene coverage of the automatic driving vehicles is improved; the scene management is responsible for managing a scene library and editing scenes in the scene library; the algorithm management is used for managing all uploaded automatic driving algorithms of the user, and the user is supported to freely load different automatic driving algorithms in different tasks in the mode; the data statistics is used for analyzing and summarizing data of all tasks, mainly focuses on accumulated virtual test mileage and abnormal scenes, and can also obtain all recorded abnormal scenes; the system configuration may configure various attributes of the entire virtual city system.

In an embodiment of the application, a user may input scene configuration parameters at a configuration interface, where the scene configuration parameters include map information and autonomous vehicle configuration information.

The map information is used for loading a test map, and after scene configuration parameters input by a user are obtained through a configuration interface, the test map used for carrying out virtual test on the automatic driving vehicle is determined according to the map information in the scene configuration parameters.

S32: and configuring the test map according to the configuration information of the automatic driving vehicles in the scene configuration parameters to generate a virtual test scene corresponding to the test map, wherein the virtual test scene comprises at least one automatic driving vehicle and at least one scene element.

Wherein the autonomous vehicle configuration information includes an autonomous driving algorithm and distribution information.

For any one of the autonomous vehicles, determining the position of the autonomous vehicle in the test map according to distribution information corresponding to the autonomous vehicle in the configuration information of the autonomous vehicle; then, generating an automatic driving vehicle configured with a corresponding automatic driving algorithm at the determined position, wherein the automatic driving algorithm is used for controlling the automatic driving vehicle to run and is equivalent to a virtual driver of the automatic driving vehicle; and randomly generating scene elements on the test map, and further generating a virtual test scene corresponding to the test map.

In an alternative embodiment, the distribution information included in the configuration information of the autonomous vehicle further includes an initial speed and a running direction of the autonomous vehicle, and the autonomous vehicle starts running according to the initial speed and the running direction after being generated on the test map.

Wherein, the scene element includes but is not limited to part or all of the following:

traffic flow vehicles, pedestrians, obstacles.

The traffic flow vehicles are traffic flow vehicles except for automatic driving vehicles, and the traffic flow vehicles run according to traffic rules in the virtual test scene.

In the present embodiment, each autonomous vehicle may be configured with a different autonomous driving algorithm, for example, different types of vehicles may be configured with different autonomous driving algorithms, such as car configured autonomous driving algorithm 1, off-road vehicle configured autonomous driving algorithm 2, truck configured autonomous driving algorithm 3, and so on.

S33: and carrying out virtual test on the automatic driving vehicle according to the virtual test scene.

S34: and determining the virtual test scene with the abnormality in the virtual test process as the virtual test scene of the automatic driving vehicle.

In the embodiment of the present application, after the virtual city system is clicked to operate based on the generated virtual test scenario, as shown in fig. 4, the virtual test is performed on the autonomous vehicle in the virtual test scenario.

Fig. 5 is a schematic diagram illustrating simulation results of a virtual test system according to an embodiment of the present application, where id (Identity document) represents an identifier of an autonomous vehicle, where id: self represents the autonomous vehicle, and id: v _100050001 represents the traffic flow vehicle; the current steering angle of the automatic driving vehicle with id self is 0 degrees, and the driving state is straight; the simulation result shows that the Speed curve, the Acceleration curve, and the Collision of the vehicle are recorded on the right side of the diagram, and the diagram shows that no Collision has occurred.

In the embodiment of the application, after the virtual test scene is detected to be abnormal, the abnormal virtual test scene is used as the virtual test scene of the automatic driving vehicle, and the virtual test scene of the automatic driving vehicle is stored in the scene library by the cloud system and can be used for future regression test.

Fig. 6A is a schematic diagram of an optional abnormal scene shown in the embodiment of the present application, which shows that an abnormal scene is a situation where an autonomous vehicle collides with a pedestrian to cause a car accident, and the abnormal scene violates a traffic law.

As shown in fig. 6B, which is a schematic diagram of another optional abnormal scenario shown in the embodiment of the present application, it can be seen that the abnormal scenario is that the autonomous vehicle 1 collides with the autonomous vehicle 2, which is also a case of violating the traffic laws and regulations.

It should be noted that the abnormal scenarios listed in the embodiments of the present application are only examples, and any abnormal scenario in the test of the autonomous vehicle is applicable to the embodiments of the present application.

In an optional implementation manner, a specific process of storing the abnormal virtual test scenario in the scenario library includes: the real-time state information of the automatic driving vehicle and/or the scene elements in the virtual test scene in a target time period is transferred into a scene file, wherein the target time period is a preset time length before the moment when the virtual test scene is abnormal; and storing the scene file into a scene library.

The real-time status information includes, but is not limited to, position, speed, acceleration, orientation information, such as heading angle, and the like.

For example, the video recording module of the virtual test system records the position, speed, acceleration, course angle, and other attribute values of the autonomous vehicle and its nearby traffic participants that affect the decision of the autonomous driving algorithm.

The nearby traffic participants refer to other autonomous vehicles in the vicinity of the autonomous vehicle, pedestrians, obstacles, traffic flow vehicles, and the like.

It should be noted that, when the scene element is a pedestrian or a traffic flow vehicle, the real-time status information of the scene element in the target time period generally includes position, speed, acceleration and course angle information; for such scene elements as obstacles, the obstacles are generally in a static state, for example, rubber road cones placed in the middle of roads, dangerous areas, road construction sections, etc. which need to temporarily separate traffic flows and guide traffic, and in such a case, since the obstacles are generally in a static state, the positions of the obstacles need to be recorded when recording real-time state information of the obstacles, and it is not necessary to record speed, acceleration, orientation information, etc. again.

In the embodiment of the present application, the preset time period may be set according to needs or experience, for example, 20 seconds or 30 seconds, and taking 30 seconds as an example, assuming that the autonomous vehicle 1 shown in fig. 6B collides with the autonomous vehicle 2 at 12:32, the target time period refers to 30 seconds before the time 12:32, that is, 12:03 to 12: 32.

The scene file comprises implementation state information of the automatic driving vehicles and/or the scene elements in the abnormal scene under each timestamp, and the implementation state information is used for restoring the scene.

For example, the data in the scene file comprises the autonomous vehicles or scene elements in the autonomous vehicles and nearby traffic participants, the positions, speeds, accelerations and heading angles of the autonomous vehicles or scene elements, which are abnormal in the scene within 0.1 second from 12:32 to 12: 32; which autonomous vehicles or scene elements are in the scene, their position, speed, acceleration and heading angle, within 0.2 seconds from 12:32 onward; which autonomous vehicles and/or scene elements are in the scene, their position, speed, acceleration, heading angle, etc. from 12:32 until 0.3 seconds ago.

In the embodiment of the application, the test roads in the virtual test scene comprise straight roads, curves, intersections, main roads, auxiliary roads and the like; the testing facilities comprise signboards, marked lines, signal lamps and the like. Therefore, each test case of the automatic driving vehicle test can be constructed in the virtual test scene, and the test cases comprise: the method comprises the steps of avoiding non-motor vehicles running reversely, avoiding non-motor vehicles crossing roads at night, avoiding non-motor vehicles running through red light, running along the non-motor vehicles in narrow roads, exceeding the front and running at a low speed, identifying emergency braking of front vehicles, passing through bus stops where buses stop, and the like.

In an optional implementation manner, if the scene configuration parameters input by the user include the working condition configuration information; configuring the test map according to the configuration information of the automatic driving vehicle and the working condition configuration information to generate a virtual test scene corresponding to the test map; specifically, when the scene elements are generated on the test map, the scene elements are generated on the test area according to the position of the automatic driving vehicle and the working condition configuration information.

The working condition configuration information refers to working conditions set for the automatic driving vehicle, and the types of the working conditions include but are not limited to pedestrian interference, acceleration and deceleration, lane changing, obstacles or signal light running and the like.

For example, if the condition configuration information includes pedestrian interference and obstacles, at least one pedestrian and at least one obstacle may be generated when the scene element is generated on the test map.

Then, the working condition of the automatic driving vehicle can be detected according to the working condition configuration information configured for the automatic driving vehicle; and when at least one automatic driving vehicle in the virtual test scene does not pass at least one working condition, determining that the virtual test scene is abnormal, and determining the abnormal virtual test scene as the virtual test scene of the automatic driving vehicle.

Assuming that the operating condition of a certain autonomous vehicle configuration includes pedestrian interference, the scenario shown in fig. 6A is taken as an example, where the scenario shown in fig. 6A indicates that the autonomous vehicle does not pass through the interference of the pedestrian but collides with the pedestrian, and thus the scenario is a virtual test scenario in which an abnormality occurs.

Referring to fig. 7, in the embodiment of the present application, after the user selects the map, the vehicle configuration, and the operating condition type of the current operation, the cloud server generates a virtual test scenario according to the map, the vehicle, the operating condition type, and the like selected by the user. And then, starting to operate the virtual city system, and recording the abnormal scene and supplementing the abnormal scene into the database after the abnormal scene occurs.

Taking the configuration interface shown in fig. 4 as an example, it can be known from the figure that the user selects all the working condition types, so that the working condition configuration information is pedestrian interference, acceleration and deceleration, lane change, obstacles and rushing to the signal lights, and the user selects 10 autonomous vehicles, the name (or identification) of the autonomous algorithm configured for the 10 autonomous vehicles is mobile-20190304, the name of the selected map is XX garden, after the virtual test scene is generated, the above 5 working conditions are set for the autonomous vehicles in the virtual test scene, and the working condition of the autonomous vehicles is detected according to the working condition configuration information configured for the autonomous vehicles; if the automatic driving vehicle passes the set working condition, the virtual test scene is normal; and if the automatic driving vehicle does not pass the set working condition, the virtual test scene is abnormal.

In an alternative embodiment, the map information may be information of a partial area in the map to be tested obtained by cutting a larger map to be tested.

For example, when the virtual city system needs to load a small city, the map information may be the map information of the whole city; when a larger city needs to be loaded, the map information may be information of one region in the city, and assuming that the a city includes 13 regions a 1-a 13, the map information may be the a1 region of the a city.

The map cutting can be applied to a scene that, for example, a user only lives in four or five blocks in a certain city, and an area 50 kilometers away from the living area of the user is meaningless, so that the map to be tested can be cut, and the cutting can be performed according to a preset rule.

As shown in fig. 8, for example, according to the square cutting, every 200m by 200m area is divided into one test map, and the test map is divided into four parts, namely, four test maps, which are A, B, C, D respectively.

In an optional implementation manner, if the test map includes a plurality of test maps, for example, as shown in fig. 8, different processing nodes need to be called to perform virtual test on a virtual test scenario corresponding to each test map, for any processing node, if it is determined that communication messages received by the processing nodes and sent by other processing nodes need to be merged, the virtual test scenario is adjusted according to a processing result after the communication messages are merged, and frames between the processing nodes are synchronized.

For example, fig. 9A and 9B show virtual test scenarios corresponding to two test maps in a fifth frame after the virtual city system starts to operate, and fig. 10A and 10B show virtual test scenarios corresponding to two test maps in the fifth frame. Wherein, the processing node corresponding to the test map a in fig. 9A is a, the processing node corresponding to the test map B in fig. 9B is B, and the processing nodes are different; it is assumed that the black boxes in these figures represent traffic flow vehicles, the white boxes represent autonomous vehicles, the triangles represent obstacles, and the circles represent pedestrians, wherein the autonomous vehicles in the circle in fig. 9A appear in fig. 10B, that is, one traffic flow vehicle in the map (the fifth frame) in fig. 9A travels into the map shown in fig. 10B from fig. 10A at the time of the sixth frame, so that the traffic flow vehicle is absent in fig. 10A in the image of the sixth frame, and the traffic flow vehicle is present in fig. 10B, in this case, two test maps have interaction of inflow and outflow of traffic flow vehicles, and thus it is determined that the communication message received by the processing node B and sent by the processing node a needs to be combined with the communication message of the processing node B, and the result obtained after the processing is shown in fig. 10B.

In an alternative embodiment, when there is traffic flow vehicle inflow and outflow interaction between multiple test maps, for example, traffic flow vehicles flow into the test map a in the test map B, C, D shown in fig. 8, where the processing node corresponding to the test map C and the test map D is the processing node C, the processing node a needs to combine the communication messages sent by the processing node B and the processing node C with the communication message of the processing node a.

Optionally, if there is inflow and outflow of traffic flow vehicles between the test map C and the test map D, similarly, the processing node C may also adjust the virtual test scene according to vehicle interaction between the two test maps.

In the embodiment of the application, all processing nodes synchronously operate, and the purpose of carrying out message communication in real time is to update frame images in real time according to communication results, ensure the accuracy of calculation results and ensure the smoothness of the images; since the entire virtual city system is distributed to run on tens or hundreds of servers. Each server may run multiple processing nodes. After the map is divided, each processing node only needs to calculate the map with a specific size. At present, a map is cut into squares and distributed to all processing nodes, and because all the processing nodes have vehicle inflow and outflow interaction, the accuracy of a calculation result can be ensured through frame synchronization among the processing nodes.

In an alternative embodiment, the input trigger intervals are acquired through the configuration interface, and the set number of traffic flow vehicles are added in the virtual test scene at intervals of the trigger intervals, wherein the traffic flow vehicles run according to the traffic rules.

The "trigger interval" shown in fig. 4 is 3s, which means that a set number of traffic flow vehicles are added to the virtual test scenario every 3 seconds, for example, one vehicle is added every 3 seconds.

Wherein, the user can also adjust the length of the trigger interval by "-" or "+" in fig. 4.

In an optional implementation manner, if the scene element includes a traffic flow vehicle, the input aggressiveness parameter may be obtained through the configuration interface, and the driving state of the traffic flow vehicle is adjusted according to the aggressiveness parameter.

The "traffic flow acceleration degree" shown in fig. 4 is 0.5, and the acceleration degree parameter can be adjusted by adjusting the progress bar of the traffic flow acceleration degree shown in fig. 4, thereby adjusting the driving state of the traffic flow vehicle.

The degree of excitement of the traffic flow refers to a driving style of a driver model of a vehicle in the traffic flow. The larger the aggressive degree parameter is, the higher the aggressive degree is, the more powerful the driving style of the driver model is, and the more serious lane changing and jamming are.

In addition, two common ways of constructing a test scenario exist at present, one of which is to generate the test scenario based on the road acquisition data playback and use the road test data acquired by the fleet, such as rossbag and Ibeo data. The data can be played back directly to reproduce the scene. And the other method is to generate a test scene based on a Monte Carlo method, and set parameters of elements in the scene by using the Monte Carlo method. A large batch of scene files may be automatically generated.

The data collected by the fleet is used when the test scene is generated based on road sampling data playback, a large amount of data can be easily obtained, and element behaviors in the scene are real. The method has the disadvantages that the collected data has no truth value and depends on manual labeling or perception methods to identify the attributes of the scene elements. And the tracks of elements in the scene are fixed, only open-loop simulation can be performed, and the data utilization rate is very low.

When the test scene is generated based on the Monte Carlo method, the parameters of the elements in the scene can be set by using the Monte Carlo method, but because the parameters are generated completely randomly, the interaction behavior among the elements is not available or unreasonable in most generated scenes. Resulting in very low scene availability.

In the embodiment of the present application, assuming that 1000 autonomous vehicles and 100 ten thousand traffic flow vehicles are generated in the map in total, the virtual city system is operated for 5 months 24 hours a day, and thousands to ten thousand abnormal scenes can be generated.

Due to the fact that the cloud simulation framework is based on the cloud simulation framework, the map of the large area can be automatically loaded, the large area simulation can be run, traffic flow vehicles, pedestrians and obstacles are automatically and randomly distributed in the area, and all the traffic flow vehicles run according to traffic rules. The vehicle to be tested can randomly run in the virtual area, and can meet various traffic scenes. Converting the scenes which cannot be processed by the automatic driving algorithm from the cloud into standard scene files, and storing the standard scene files into a scene library for future regression testing; in addition, all traffic flow vehicles in the embodiment of the application run according to traffic rules, vehicles to be detected randomly run in a virtual area, interaction behaviors among elements in a generated scene are reasonable, and the scene availability is improved.

Referring to fig. 11, a schematic diagram of a timing sequence for implementing a method for determining a virtual test scenario of an autonomous vehicle is shown. The specific implementation flow of the method is as follows:

step 1100: the cloud server determines a map, vehicle configuration and working condition type selected by a user to run at the time;

step 1101: the cloud server loads the requested map from the map server, cuts the map and sends the map to different processing nodes;

step 1102: after calling different processing nodes to load a small map, the cloud server automatically generates at least one automatic driving vehicle carrying an automatic driving algorithm and scene elements on the small map;

step 1103: the frame synchronization of all processing nodes is carried out in the cloud server, so that the operation and the message communication of all the processing nodes are ensured to be synchronous;

step 1104: the cloud server continuously records data for 30 seconds;

step 1105: the cloud server judges whether the virtual test scene is abnormal or not, if so, the step 1106 is executed, otherwise, the step 1103 is returned;

step 1106, after the abnormal scene occurs, the cloud server automatically stores the 30 second data record into a scene file, and stores the scene file into a background scene database.

As shown in fig. 12, a schematic structural diagram of an apparatus for determining a virtual test scenario of an autonomous vehicle may include:

a parameter obtaining unit 1200, configured to obtain an input scene configuration parameter, and determine a test map for performing a virtual test on an autonomous vehicle according to map information in the scene configuration parameter;

a scene generating unit 1201, configured to configure the test map according to the configuration information of the autonomous driving vehicle in the scene configuration parameters to generate a virtual test scene corresponding to the test map, where the virtual test scene includes at least one autonomous driving vehicle and at least one scene element;

a virtual test unit 1202, configured to perform a virtual test on the autonomous vehicle according to the virtual test scenario; and determining the virtual test scene with the abnormality in the virtual test process as the virtual test scene of the automatic driving vehicle.

Optionally, the apparatus further includes a scene storage unit 1203:

the scene storage unit 1203 is configured to store real-time state information of the determined autonomous vehicle and/or scene elements in a virtual test scene of the autonomous vehicle as a scene file within a target time period, where the target time period is a preset time length before a time when the virtual test scene is abnormal; and storing the scene file into the scene library.

Optionally, the scene generation unit 1201 is specifically configured to:

for any one automatic driving vehicle, determining the position of the automatic driving vehicle in the test map according to distribution information corresponding to the automatic driving vehicle in the automatic driving vehicle configuration information;

generating an automatic driving vehicle configured with a corresponding automatic driving algorithm at a determined position according to an automatic driving algorithm corresponding to the automatic driving vehicle in the automatic driving vehicle configuration information, wherein the automatic driving algorithm is used for controlling the automatic driving vehicle to run; and

and generating scene elements on the test map so as to generate a virtual test scene corresponding to the test map.

Optionally, the test map includes a plurality of test maps, and the virtual test unit 1202 is further configured to:

if different processing nodes are called to respectively carry out virtual testing on the virtual test scene corresponding to each test map, communication messages received by the processing nodes and sent by other processing nodes need to be combined, the virtual test scene is adjusted according to a processing result after the communication messages are combined, and frames among the processing nodes are synchronous;

and carrying out virtual test on the automatic driving vehicle according to the adjusted virtual test scene.

Optionally, the scene configuration parameter includes operating condition configuration information;

the scene generation unit 1201 is specifically configured to:

configuring the test map according to the configuration information of the automatic driving vehicle and the working condition configuration information to generate a virtual test scene corresponding to the test map;

when generating a scene element on the test map, the scene generation unit 1201 is specifically configured to:

and generating scene elements on the test area according to the position of the automatic driving vehicle and the working condition configuration information so as to generate a virtual test scene corresponding to the test map.

Optionally, the virtual test scenario with the exception includes:

Optionally, the parameter obtaining unit 1200 is further configured to:

For convenience of description, the above parts are separately described as modules (or units) according to functional division. Of course, the functionality of the various modules (or units) may be implemented in the same one or more pieces of software or hardware when implementing the present application.

Having described the method and apparatus for determining a virtual test scenario for an autonomous vehicle in accordance with an exemplary embodiment of the present application, an apparatus for determining a virtual test scenario for an autonomous vehicle in accordance with another exemplary embodiment of the present application is described.

As will be appreciated by one skilled in the art, aspects of the present application may be embodied as a system, method or program product. Accordingly, various aspects of the present application may be embodied in the form of: an entirely hardware embodiment, an entirely software embodiment (including firmware, microcode, etc.) or an embodiment combining hardware and software aspects that may all generally be referred to herein as a "circuit," module "or" system.

In some possible implementations, a computing device according to the present application may include at least one processor, and at least one memory. Wherein the memory stores program code which, when executed by the processor, causes the processor to perform the steps of the method of determining a virtual test scenario for an autonomous vehicle according to various exemplary embodiments of the present application described above in this specification. For example, the processor may perform the steps as shown in fig. 3.

The device for determining the virtual test scenario of the autonomous vehicle according to this embodiment has a similar structure to the device for determining the virtual test scenario of the autonomous vehicle shown in fig. 12, and is not repeated here.

The computing device 130 according to this embodiment of the present application is described below with reference to fig. 13. The computing device 130 of fig. 13 is only an example and should not bring any limitations to the functionality or scope of use of the embodiments of the present application.

As shown in fig. 13, computing device 130 is embodied in the form of a general purpose computing device. Components of computing device 130 may include, but are not limited to: the at least one processing unit 131, the at least one memory unit 132, and a bus 133 connecting various system components (including the memory unit 132 and the processing unit 131).

Bus 133 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a processor, or a local bus using any of a variety of bus architectures.

The storage unit 132 may include readable media in the form of volatile memory, such as Random Access Memory (RAM)1321 and/or cache memory 1322, and may further include Read Only Memory (ROM) 1323.

Storage unit 132 may also include a program/utility 1325 having a set (at least one) of program modules 1324, such program modules 1324 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

Computing device 130 may also communicate with one or more external devices 134 (e.g., keyboard, pointing device, etc.), with one or more devices that enable a user to interact with computing device 130, and/or with any devices (e.g., router, modem, etc.) that enable computing device 130 to communicate with one or more other computing devices. Such communication may occur via input/output (I/O) interfaces 135. Also, computing device 130 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the internet) via network adapter 136. As shown, network adapter 136 communicates with other modules for computing device 130 over bus 133. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with computing device 130, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

In some possible embodiments, the aspects of the method for determining a virtual test scenario for an autonomous vehicle provided by the present application may also be implemented in the form of a program product comprising program code for causing a computer device to perform the steps of the method for determining a virtual test scenario for an autonomous vehicle according to various exemplary embodiments of the present application described above in this specification, when the program product is run on a computer device, for example, the computer device may perform the steps as shown in fig. 3.

The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The scenario-structured program product of the embodiments of the present application may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a computing device. However, the program product of the present application is not limited thereto, and in this document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with a command execution system, apparatus, or device.

A readable signal medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable signal medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with a command execution system, apparatus, or device.

Program code embodied on a readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Program code for carrying out operations of the present application may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user computing device, partly on the user equipment, as a stand-alone software package, partly on the user computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).

It should be noted that although in the above detailed description several units or units of the device are mentioned, such cutting is merely exemplary and not mandatory. Indeed, the features and functions of two or more units described above may be embodied in one unit, according to embodiments of the application. Conversely, features and functions of one unit described above may be further cut into embodiments by a plurality of units.

Further, while the operations of the methods of the present application are depicted in the drawings in a particular order, this does not require or imply that these operations must be performed in this particular order, or that all of the illustrated operations must be performed, to achieve desirable results. Additionally or alternatively, certain steps may be omitted, multiple steps combined into one step execution, and/or one step broken down into multiple step executions.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims

1. A method of determining a virtual test scenario for an autonomous vehicle, the method comprising:

2. The method of claim 1, wherein the method further comprises:

the method comprises the steps that real-time state information of the determined automatic driving vehicles and/or scene elements in a virtual test scene of the automatic driving vehicles in a target time period is transferred and stored as a scene file, wherein the target time period is a preset time length before the virtual test scene is abnormal;

and storing the scene file into the scene library.

3. The method of claim 1, wherein the configuring the test map according to the configuration information of the autonomous driving vehicle in the scenario configuration parameters to generate a virtual test scenario corresponding to the test map comprises:

4. The method of claim 1, wherein the test map includes a plurality of test maps, and if different processing nodes are called to perform virtual testing on the virtual test scenario corresponding to each test map, the method further includes, for any one of the processing nodes:

if the communication messages received by the processing nodes and sent by other processing nodes need to be merged, adjusting the virtual test scene according to the processing result after the communication messages are merged, and synchronizing frames among the processing nodes;

5. The method of claim 3, wherein the scenario configuration parameters include condition configuration information;

the configuring the test map according to the configuration information of the automatic driving vehicle in the scene configuration parameters to generate a virtual test scene corresponding to the test map includes:

the generating of the scene element on the test map comprises:

6. The method of claim 5, wherein the abnormal virtual test scenario comprises:

7. The method of claim 6, wherein the method further comprises:

8. The method of claim 7, wherein the method further comprises:

9. An apparatus for determining a virtual test scenario for an autonomous vehicle, comprising:

the parameter determining unit is used for acquiring input scene configuration parameters and determining a test map for carrying out virtual test on the automatic driving vehicle according to map information in the scene configuration parameters;

the virtual test unit is used for carrying out virtual test on the automatic driving vehicle according to the virtual test scene; and are

10. The apparatus of claim 9, wherein the apparatus further comprises a scene storage unit:

the scene storage unit is used for storing real-time state information of the determined automatic driving vehicles and/or scene elements in a virtual test scene of the automatic driving vehicles as a scene file in a target time period, wherein the target time period is a preset time length before the moment when the virtual test scene is abnormal; and storing the scene file into the scene library.

11. The apparatus of claim 9, wherein the scene generation unit is specifically configured to:

12. The apparatus of claim 9, wherein the test map comprises a plurality, the virtual test unit further to:

if different processing nodes are called to respectively perform virtual testing on the virtual test scene corresponding to each test map, aiming at any one processing node, if the communication messages received by the processing nodes and sent by other processing nodes are determined to be needed to be combined, the virtual test scene is adjusted according to the processing result after the communication messages are combined, and frames among the processing nodes are synchronous;

13. The apparatus of claim 11, wherein the scenario configuration parameter comprises a condition configuration information;

the scene generation unit is specifically configured to:

when generating a scene element on the test map, the scene generation unit is specifically configured to:

14. An electronic device, comprising a processor and a memory, wherein the memory stores program code which, when executed by the processor, causes the processor to perform the steps of the method of any of claims 1 to 8.

15. A computer-readable storage medium, characterized in that it comprises program code for causing an apparatus for determining virtual test scenarios for autonomous vehicles to carry out the steps of the method according to any one of claims 1 to 8, when said program product is run on said apparatus for determining virtual test scenarios for autonomous vehicles.