CN116932340A

CN116932340A - Scene operation method, device, computer equipment and storage medium

Info

Publication number: CN116932340A
Application number: CN202210340114.5A
Authority: CN
Inventors: 胡太群
Original assignee: Tencent Cloud Computing Beijing Co Ltd
Current assignee: Tencent Cloud Computing Beijing Co Ltd
Priority date: 2022-04-01
Filing date: 2022-04-01
Publication date: 2023-10-24

Abstract

The embodiment of the application discloses a scene operation method, a device, computer equipment and a storage medium, which can be applied to the field of automatic driving, wherein the method comprises the following steps: acquiring a scene operation requirement, and acquiring a parameter generalization rule related to the scene operation requirement, wherein the parameter generalization rule is used for determining one or more traffic participation objects to be generalized and the value ranges of corresponding scene parameters; wherein, a traffic participation object to be generalized corresponds to one or more scene parameters, and a corresponding parameter generalization rule exists for one scene parameter; performing generalization processing on corresponding scene parameters according to the acquired parameter generalization rules, and generating one or more generalized simulation scenes; and running the generated generalized simulation scene, so that scene generation efficiency can be improved.

Description

Scene operation method, device, computer equipment and storage medium

Technical Field

The present application relates to the field of computer technologies, and in particular, to a scene operation method, a device, a computer device, and a storage medium.

Background

With the continuous and deep development of computer technology, various automatic driving algorithms are developed to further improve the production life quality of users, so as to improve the efficiency and driving safety of the users in the driving field, and before the automatic driving algorithm is actually put into use, the algorithm capability of the automatic driving algorithm needs to be tested to ensure the safety when the automatic driving algorithm is adopted to control the actual vehicle to run, and when the automatic driving algorithm is tested, the automatic driving algorithm is usually deployed into a large number of test scenes to run, and the current method for obtaining a large number of test scenes is usually obtained by manually determining traffic participation objects and adjusting related scene parameters, so that the efficiency of scene generation is low due to the adoption of the current scene generation mode.

Disclosure of Invention

The embodiment of the invention provides a scene operation method, a scene operation device, computer equipment and a storage medium, which can improve scene generation efficiency.

In one aspect, an embodiment of the present invention provides a scenario running method, including:

acquiring a scene operation requirement, and acquiring a parameter generalization rule related to the scene operation requirement, wherein the parameter generalization rule is used for determining one or more traffic participation objects to be generalized and the value ranges of corresponding scene parameters; wherein, a traffic participation object to be generalized corresponds to one or more scene parameters, and a corresponding parameter generalization rule exists for one scene parameter;

performing generalization processing on corresponding scene parameters according to the acquired parameter generalization rules, and generating one or more generalized simulation scenes;

and running the generated generalized simulation scene.

In still another aspect, an embodiment of the present invention provides a scenario running apparatus, including:

the system comprises an acquisition unit, a parameter generalization unit and a processing unit, wherein the acquisition unit is used for acquiring scene operation requirements and acquiring parameter generalization rules related to the scene operation requirements, wherein the parameter generalization rules are used for determining one or more traffic participation objects to be generalization processed and the value ranges of corresponding scene parameters; wherein, a traffic participation object to be generalized corresponds to one or more scene parameters, and a corresponding parameter generalization rule exists for one scene parameter;

The processing unit is used for carrying out generalization processing on corresponding scene parameters according to the acquired parameter generalization rule and generating one or more generalized simulation scenes;

the processing unit is further used for running the generated generalized simulation scene.

In yet another aspect, an embodiment of the present invention provides a computer device, including a processor, an input device, an output device, and a memory, where the processor, the input device, the output device, and the memory are connected to each other, and the memory is configured to store a computer program that supports the computer device to perform the above method, where the computer program includes program instructions, and where the processor is configured to invoke the program instructions to perform the following steps:

And running the generated generalized simulation scene.

In yet another aspect, an embodiment of the present application provides a computer readable storage medium, where a program instruction is stored, where the program instruction, when executed by a processor, is used to perform the scenario running method according to the first aspect.

In the embodiment of the application, after the computer equipment acquires the scene operation requirement, the computer equipment can acquire the parameter generalization rule related to the scene operation requirement, and further, the scene parameters of the corresponding traffic participation objects can be subjected to generalization processing according to the traffic participation objects to be subjected to generalization processing and the value range of the corresponding scene parameters indicated by the acquired parameter generalization rule, so that one or more generalization simulation scenes are acquired, and after the computer equipment acquires one or more generalization simulation scenes, the automatic driving algorithm to be tested can be carried in a host vehicle of the generalization simulation scene, so that the test of the algorithm performance of the automatic driving algorithm is realized. The method has the advantages that the computer equipment generates the generalization simulation scene based on the requirement in real time after the requirement of scene operation is acquired, the generation of the generalization simulation scene is not needed in advance by the computer equipment, further, the storage pressure of the computer equipment on the generalization simulation scene generated in advance can be reduced, the effective release of storage resources of the computer equipment is realized, the generation efficiency of massive generalized simulation scenes can be improved based on the release of the storage resources, in addition, if the traffic participation objects to be subjected to the generalization treatment indicated by the parameter generalization rule adopted by the computer equipment when the generation of the generalization simulation scene is performed are mostly identical, the computer equipment can generate more similar generalized simulation scenes, and when the automatic driving algorithm is tested by adopting the more similar generalized simulation scene, the accuracy of the computer equipment on the test of the algorithm boundary capability of the automatic driving algorithm can be improved, and when the difference of the generalized simulation scenes generated by the computer equipment based on the parameter generalization rule is larger, the automatic driving algorithm can be tested by adopting different computer equipment, and further, the automatic driving algorithm can be tested comprehensively.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic diagram of a traffic simulation scenario provided by an embodiment of the present invention;

FIG. 2 is a schematic flow chart of a scenario running method provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram of a scenario running method according to an embodiment of the present invention;

FIG. 4 is a schematic flow chart of a scenario running method provided by an embodiment of the present invention;

FIG. 5a is a schematic diagram of a loading display of a template scene provided by an embodiment of the present invention;

FIG. 5b is a schematic diagram of a parameter generalization interface provided by an embodiment of the present invention;

FIG. 5c is a schematic diagram of generating and running a generalized simulation scenario provided by an embodiment of the present invention;

FIG. 5d is a schematic diagram of a query interface provided by an embodiment of the present invention;

FIG. 6 is a schematic block diagram of a scenario running apparatus provided in an embodiment of the present invention;

Fig. 7 is a schematic block diagram of a computer device provided by an embodiment of the present application.

Detailed Description

The embodiment of the application provides a scene operation method, which enables computer equipment to perform generalization processing on corresponding scene parameters of corresponding traffic participation objects by acquiring parameter generalization rules related to scene operation requirements when the computer equipment detects that the operation requirements of the scenes exist, thereby realizing the generation and operation of one or more generalized simulation scenes, enabling the computer equipment to perform the generation and operation process of the corresponding generalized simulation scenes based on the actual scene operation requirements without locally storing massive simulation scenes (namely scene files), thereby realizing the effective saving of storage resources of the computer equipment and avoiding the waste of storage space of the computer equipment. In one embodiment, since the algorithm test process is usually implemented by running a corresponding test scene in a simulation system and carrying the automatic driving algorithm in a host vehicle (egoco) of the test scene, the computer device can determine that a requirement for performing scene operation exists when determining that the automatic driving algorithm needs to be subjected to the algorithm test, where the simulation system refers to a traffic simulation system of an automatic driving automobile, the test scene may also be called a simulation scene or a traffic simulation scene, etc., specifically refers to a simulated traffic scene constructed in the simulation system and used for simulating a real traffic situation, at least includes a host vehicle (egoco) carrying the automatic driving algorithm in the test scene, and in addition, the test scene may also include other traffic participation objects (or traffic participants), such as dynamic or static objects including participation interaction in the test scene, and specifically may be vehicles, pedestrians, animals or obstacles, etc. In one embodiment, the test scenario may be as specifically shown in FIG. 1, and the host vehicle in the test scenario may be, for example, a vehicle as indicated at 10 in FIG. 1. While other traffic participants of the test scene may be, for example, vehicles or pedestrians, etc., in addition to the 10-labeled vehicles in fig. 1. In one embodiment, the computer device may be a device on which the simulation system is installed, or may be a related device on which the simulation system is installed, which is not limited in the embodiment of the present application. The test scenario running in the simulation system is generally a custom or XML format text file (a standard format file of a test scenario) based on the OpenScenario standard, that is, the test scenario referred to in the embodiment of the present application is a scenario file (or description file) corresponding to the test scenario, where the scenario file (or description file) is used to describe the position of each traffic participant object in the corresponding test scenario, the driving state in the corresponding test scenario, and trigger information for triggering the driving state switching of the traffic participant object.

In one embodiment, the computer device may select one or more test scenes from existing test scenes in the traffic simulation system as a template scene, where the template scene refers to a test scene that provides a sample template function for scene generalization, and is also a specific test scene, and then the computer device may perform generalization configuration processing on each scene parameter in any one of the template scenes, so as to obtain a parameter generalization rule corresponding to any one of the scene parameters in each of the template scenes, that is, the computer device may obtain parameter generalization rules corresponding to different scene parameters of different traffic participation objects, and after the computer obtains the parameter generalization rules corresponding to different scene parameters, the obtained parameter generalization rules may be stored. After the computer device stores the obtained parameter generalization parameters, the computer device can acquire parameter generalization rules related to scene operation requirements when determining that the scene operation requirements exist, wherein the related parameter generalization rules related to the scene operation requirements refer to: if the scene operation requirement is used for indicating the total number of the generalized simulation scenes to be obtained, the correlation with the scene operation requirement is equal to the total number of the generalized simulation scenes to be obtained; and if the scene operation requirement is used for indicating a template scene used for generalization in one or more template scenes, the template scene to be subjected to generalization is related to the scene operation requirement. Under the general condition, when the stored parameter generalization rules are obtained by carrying out parameter generalization processing on scene parameters in a template scene, the related parameter generalization rules obtained by the computer equipment based on the scene operation requirement are the parameter generalization rules which are equal to the total number of scenes of the to-be-obtained generalization simulation scene; when the stored parameter generalization rules are obtained by generalizing scene parameters in different template scenes, the related parameter generalization rules acquired by the computer equipment based on the scene operation requirements are the parameter generalization rules corresponding to all scene parameters in each template scene to be subjected to the generalization processing.

Based on the obtained parameter generalization rule related to the scene operation requirement, the computer device can perform generalization processing on the corresponding scene parameters based on the parameter generalization rule, and obtain one or more generalization simulation scenes, in one embodiment, the parameter generalization rule is used for indicating the value range of the corresponding scene parameters of one or more traffic participation objects to be generalized, so that it can be understood that the generalization processing refers to the generation of a large number of one or more generalization simulation scenes similar to a sample scene in batches according to a certain arrangement and combination rule by using the parameters of each traffic participation object in a certain sample test scene (i.e. a template scene), and therefore, when the computer device performs generalization processing based on the parameter generalization rule, namely, the value range indicated by the parameter generalization rule, the computer device performs random value processing on the corresponding scene parameters, and performs random combination on the different scene parameters after the value is taken, so as to generate the corresponding generalization simulation scene.

In a specific implementation, if the parameter generalization rule obtained by the computer device is used for indicating the value ranges of corresponding scene parameters corresponding to a traffic participation object a and a traffic participation object b to be subjected to generalization processing, and if the scene parameters corresponding to the traffic participation object a include a scene parameter 1 and a scene parameter 2 and the scene parameters corresponding to the traffic participation object b include a scene parameter 3, a scene parameter 4 and a scene parameter 5, the parameter generalization rule obtained by the computer device includes the parameter generalization rule corresponding to each of the scene parameters 1 to 5, and after the computer device obtains the parameter generalization rule corresponding to each scene parameter, any parameter generalization rule can be adopted to randomly value the corresponding scene parameter, and randomly combining the randomly valued scene parameters, thereby realizing the generation of the generalization simulation scene. In one embodiment, the random value and combination processing are performed on the corresponding scene parameters based on the parameter generalization rule, so that the generation of the generalization simulation scene is realized, and the generated generalization simulation scene generated by the computer equipment is similar to the template scene because the traffic participation object contained in the generated generalization simulation scene is the same as the traffic participation object indicated by the acquired parameter generalization rule, and the parameter generalization rule is obtained by performing the generalization configuration processing on the template scene. And based on the operation process of the generalized simulation scene, even if the automatic driving algorithm is subjected to the algorithm test process, the automatic driving algorithm is operated by generating the simulation scene in real time, and compared with the mode of operating the automatic driving algorithm by adopting the simulation scene stored in advance, the method can effectively lighten the storage pressure of the computer equipment on the simulation scene, thereby realizing the effective saving of the computing resources of the computer equipment.

In one embodiment, after the computer device runs the generated generalized simulation scenes, the computer device may obtain the running result of each generalized simulation scene, store the corresponding running result, delete the generated generalized simulation scenes, and re-generate the generalized simulation scenes according to the generalization requirements when the test of the automatic driving algorithm is needed next time. The method comprises the steps that a generalized simulation scene contains more scene description information, so that more memory space of the computer equipment is consumed by the storage of the generalized simulation scene, the computer equipment does not need to spend more memory resources for storing the generalized simulation scene after the generalized simulation scene is generated and operated, only the operation result of the generalized simulation scene is stored, the evaluation result of an automatic driving algorithm can be obtained from the stored operation result, the minimum occupation of the computing resources of the computer equipment is effectively realized, and the computing resources of the computer equipment are effectively promoted.

Referring to fig. 2, a schematic flowchart of a scenario operation method according to an embodiment of the present application may be executed by the above-mentioned computer device, and as shown in fig. 2, the method may include:

S201, acquiring scene operation requirements, and acquiring parameter generalization rules related to the scene operation requirements, wherein the parameter generalization rules are used for determining one or more traffic participation objects to be generalized and the value ranges of corresponding scene parameters.

S202, performing generalization processing on corresponding scene parameters according to the acquired parameter generalization rules, and generating one or more generalized simulation scenes.

In step S201 and step S202, the computer device may obtain a parameter generalization rule related to the scene operation requirement from the stored generalization parameters, where the stored parameter generalization rule may be obtained by the computer device after performing parameter generalization processing on the scene parameters of each traffic participant in the template scene, or the parameter generalization rule may be preset in advance by the computer device based on the test requirement for the autopilot algorithm. In one embodiment, the process of presetting and storing one or more parameter generalization rules in advance based on the test requirement of the autopilot algorithm is that the computer device can firstly determine traffic participation objects required by the test of the autopilot algorithm and scene parameters involved in the test process of each traffic participation object, and further, the computer device can set a corresponding parameter value range for the determined scene parameters of each traffic participation object, so that the parameter value range corresponding to the scene parameters of each traffic participation object set for the autopilot algorithm can be stored as the parameter generalization rule. In a specific implementation, if the computer device determines that the traffic participation object required for performing the test of the autopilot algorithm includes a host vehicle a and a traffic vehicle b, and the scene parameters set for the host vehicle a include a speed, an acceleration, and a collision time, and the scene parameters set for the traffic vehicle b include a speed and an acceleration, then the computer device may store the parameter value ranges corresponding to each of the scene parameters of the host vehicle a and the traffic vehicle b as the parameter generalization rule by setting a corresponding parameter value range for the speed, the acceleration, and the collision time included in the scene parameters of the host vehicle a, and setting a corresponding parameter value range for the speed and the acceleration included in the scene parameters of the traffic vehicle b, respectively.

In another implementation manner, the computer device may acquire a template scene first, where the template scene acquired by the computer device includes one or more traffic parameter objects and parameter values of scene parameters of each traffic participation object, and after acquiring the template scene, the computer device may use the traffic participation object included in the template scene as a traffic participation object required when performing the autopilot algorithm, and further may perform parameter generalization processing on the traffic parameter that has been valued by each traffic participation object in the template scene, to obtain a parameter generalization interval of the scene parameter of each traffic participation object in the template scene, and further the computer device may store the parameter generalization interval of the scene parameter of each traffic participation object in the template scene as the parameter generalization rule. Wherein, a traffic participation object to be generalized corresponds to one or more scene parameters, and a scene parameter has a corresponding parameter generalization rule. In the embodiment of the present application, the stored parameter generalization rule is mainly used to describe the situation obtained by performing parameter generalization processing on each scene parameter in the template scene, and when the stored parameter generalization rule is obtained by storing in other manners, the embodiment of the present application can be seen.

In one embodiment, the number of the traffic participation objects required for the automatic driving algorithm test determined by the computer device may be one or more, if the stored parameter generalization rule is obtained by setting a corresponding parameter value range for each scene parameter of the traffic participation objects required for the automatic driving algorithm test, the computer device may first obtain the traffic participation object required for the automatic driving algorithm to be tested when obtaining the parameter generalization rule related to the scene operation requirement according to the obtained scene operation requirement, and the obtained traffic participation object required for the automatic driving algorithm to be tested may be part or all of the traffic participation object corresponding to the stored parameter generalization rule, and then the computer device may use the obtained parameter generalization rule of the traffic participation object corresponding to the scene parameter required for the automatic driving algorithm to be tested when performing the algorithm test as the parameter generalization rule related to the scene operation requirement.

In one embodiment, if the parameter generalization rule stored in the computer device is obtained by performing parameter generalization processing on the scene parameters of each traffic participation object in the template scene, when the computer device obtains the parameter generalization rule related to the scene operation requirement, part or all of the stored parameter generalization rule can be used as the parameter generalization rule related to the scene operation requirement based on the indication of the scene operation requirement.

After the computer equipment obtains the corresponding parameter generalization rule, the computer equipment can carry out generalization processing on the corresponding scene parameters according to the obtained parameter generalization rule and generate one or more generalization simulation scenes, in the specific implementation, when the computer equipment carries out the generalization processing on the corresponding scene parameters according to the obtained parameter generalization rule, the computer equipment can carry out random value processing on the basis of the value range of the scene parameters corresponding to the traffic participation object to be subjected to the generalization processing indicated by the parameter generalization rule, and after the random value processing, the random value processing is carried out on the scene parameters after the random value processing according to the corresponding relation between the traffic participation object to be subjected to the generalization processing and the scene parameters, thereby carrying out random combination on the scene parameters after the random value processing so as to generate one or more generalization simulation scenes, wherein the obtained generalization simulation scenes comprise one or more traffic participation scenesThe object and the corresponding scene parameter are valued. For example, if the obtained parameter generalization rule includes a parameter generalization rule corresponding to a speed of the host vehicle a (one scene parameter of the host vehicle a) of 100-199 meters per second (m/s), and a parameter generalization rule corresponding to an acceleration of the host vehicle a (another scene parameter of the host vehicle a) of 1-9 meters per square second (m/s) ² ) In addition, the obtained parameter generalization rule further includes a parameter generalization rule corresponding to the speed of the traffic vehicle b (a scene parameter of the host vehicle b) of 80-179m/s, and then the computer device performs the generalization processing based on the obtained parameter generalization rule to obtain a generalization simulation scene which may be 120m/s of the speed of the host vehicle a and 2m/s of the acceleration ² The speed of the host vehicle b is 100m/s.

In one embodiment, when the computer device performs random value-taking on the parameter-generalization rule, if a value-taking interval of random value-taking is specified, based on a value-taking interval specified by the parameter-generalization rule, the computer device may determine, based on the obtained parameter-generalization rule, a total number of generalization simulation scenes generated by the obtained parameter-generalization rule and supported by the generalization process, where, as can be seen in the above example, if the value-taking interval of random value-taking is specified as 1, then the number of generalization simulation scenes generated by the computer device supported by the parameter-generalization rule shown in the above example is 100×10×100=100000. It should be noted that, the value intervals set by the computer device for the parameter generalization rules of different scene parameters may be the same or different, which is not limited by the embodiment of the present application. After generating one or more generalized simulation scenes, the computer device may run the generated generalized simulation scenes, i.e. may execute step S203 instead.

S203, running the generated generalized simulation scene.

In one embodiment, the process of running the generated generalized simulation scenes by the computer device is that an automatic driving algorithm is carried in a host vehicle of each generalized simulation scene, and the automatic driving algorithm is adopted to drive the host vehicle to run in the corresponding generalized simulation scene, so that it can be understood that a test result aiming at the automatic driving algorithm can be obtained based on the running of the computer device on the generalized simulation scenes.

In one embodiment, after the computer device obtains the operation result of each generalized simulation scene, the obtained operation result may be stored, so that the algorithm performance of the automatic driving algorithm is analyzed based on the obtained operation result, and after the computer device stores the operation result of the corresponding generalized simulation scene, the corresponding generalized simulation scene may be deleted, so as to avoid storage resource waste caused by the computer device based on the storage of the corresponding generalized simulation scene, thereby effectively saving storage resources of the computer device. Based on the storage of the operation results of the computer equipment on the corresponding simulation scenes, the computer equipment can realize the effective analysis of the automatic driving algorithm based on the stored operation results, and the accuracy and the effectiveness of the algorithm analysis results of the automatic driving algorithm are ensured under the condition of effectively releasing the storage resources of the computer equipment.

When the computer device executes the scene operation method, the computer device mainly relates to a scene loading module, a generalization parameter configuration module, a generalization parameter storage module, a generalization scene generation module, a generalization scene operation module, a generalization scene query module and a generalization scene derivation module, wherein the overall process of performing scene operation by the computer device based on the module functions of the modules can be as shown in fig. 3, firstly, the scene loading module loads a selected template scene, then the generalization parameter configuration module performs generalization configuration processing on related scene parameters of each traffic participation object to obtain parameter generalization rules corresponding to each scene parameter, and the parameter generalization rules of each scene parameter are stored in a database. And after the generalization scene operation module executes the generalization scene, discarding the generated generalization scene file, and only reserving an execution result and storing the execution result into a database.

In addition, the related objects can input query information, wherein the query information can be scene parameters, scene labels and the like of the traffic participation objects, and the generalized scene query module can search the operation results of the corresponding generalized simulation scenes in the database according to the query information. And when the operation result is obtained, the corresponding operation result is displayed, if the operation result does not exist, the scene information can be generated according to the output scene parameters, and the scene information comprises template scene information (scene name, scene label, scene type, creator, high-precision map name and the like), generalization parameter information and the like.

The following describes in detail the implementation of the scenario execution method with reference to fig. 4 and relevant modules of the computer device. As shown in fig. 4, the method may include:

s401, a template scene is acquired, wherein the template scene comprises one or more traffic participation objects.

S402, performing generalization configuration processing on scene parameters of each traffic participation object to obtain parameter generalization rules corresponding to each scene parameter of each traffic participation object.

S403, generating generalization information data according to parameter generalization rules corresponding to each scene parameter of each traffic participation object in the template scene, and storing the generalization information data into a database.

In step S401 to step S403, the template scene acquired by the computer device is a part or all of the simulation scenes selected by the operation object from the one or more standard simulation scenes stored in the cloud platform, that is, the one or more standard simulation scenes are stored in the cloud platform, the operation object may select one or more template simulation scenes (or template scenes) from the one or more standard simulation scenes stored in the cloud platform, and send the selected template simulation scenes to the computer device through the cloud platform, so that the computer device acquires the one or more template scenes. The operation object may be a consumer object on the consumer side requesting for evaluation of the autopilot algorithm, or may be a producer object on the producer side performing generation of the generalized simulation scene, where in the embodiment of the present application, the object type of the operation object is not limited. In one embodiment, the cloud platform for supporting the operation object to perform template scene selection is implemented based on a cloud technology, and in the cloud platform, storage of a standard simulation scene can also be implemented by adopting a blockchain technology to ensure accuracy of the standard simulation scene stored in the cloud platform, wherein the standard simulation scene refers to a description file for performing scene description by adopting a standard file description format.

After the computer device obtains the template simulation scene, a scene loading module in the computer device can select a corresponding template scene from the one or more obtained template scenes to load and display, wherein the interface after loading and displaying a certain template scene can be shown in fig. 5a, as shown in fig. 5a, based on the loading and displaying the corresponding template scene by the scene loading module, an operation object can visually view traffic participation objects contained in the template scene based on the displayed scene interface as shown in fig. 5a, such as the contained traffic participation objects comprise running vehicles named as a host vehicle, car1 (vehicle 1) and car2 (vehicle 2), lane lines and the like, and attribute information of each traffic participation object in the template scene can be visually checked, for example, the speed and model of the running vehicle. In one embodiment, based on the loading display of a certain template scene by the scene loading module, the operation object can view the displayed template scene and execute corresponding editing operation in the displayed template scene, so as to edit the scene parameters contained in the displayed template scene, wherein the scene interface displayed by the computer device contains a scene generalization component, which may be a key marked by 40 in fig. 5a, for example, then the computer device can support to acquire the operation executed by the operation object in the displayed scene interface after determining that the scene generalization component is selected, and take the acquired operation as the editing operation for the target scene, so as to edit the scene parameters contained in the template scene.

In one embodiment, when the computer device detects that the scene generalization component is selected, a parameter generalization interface may be displayed, so that an operation object may perform configuration processing on corresponding scene parameters of a displayed template scene in the parameter generalization interface, thereby obtaining a parameter generalization rule corresponding to each scene parameter, where a process of displaying the parameter generalization interface by the computer device after the scene generalization component is selected may be as shown in fig. 5b, and an exemplary parameter generalization interface displayed by the computer device may be an interface labeled by 401 in fig. 5 b. The scene name of the template scene included in the parameter generalization interface displayed by the computer device is base_abc_001, and the traffic participation object included in the template scene named base_abc_001 includes: after displaying the parameter generalization interfaces marked 401 in fig. 5b, the computer device may perform the generalization configuration processing on the scene parameters related to each traffic participant in the template scene named base_abc_001 based on the parameter generalization interfaces, that is, the generalization parameter configuration module may perform the generalization configuration processing on the scene parameters corresponding to the host vehicle named planner0 and the traffic vehicles named car1 and car2 based on the parameter generalization interfaces, thereby obtaining the parameter generalization rule of each scene parameter of each scene traffic participant.

Based on the displayed parameter generalization interface, the computer device displays the parameter generalization interface and realizes the process of generalizing and configuring the scene parameters of any traffic participation object in the displayed template scene, which comprises the following steps: the computer device first determines a current configuration object from one or more traffic participant objects contained in the template scene and outputs one or more scene parameters of the current configuration object. Based on the parameter generalization interface denoted by 401 in fig. 5b, the traffic vehicle denoted by car1 in the template scene denoted by base_abc_001 is selected as the current configuration object, and therefore, the computer device may further output one or more scene parameters of the traffic vehicle denoted by car1, thereby acquiring a generalization configuration process performed on the corresponding scene parameters, wherein the scene parameters corresponding to the current configuration object output and displayed by the computer device are parameters in which the current configuration object can be subjected to the generalization configuration process, and three parameters capable of the generalization configuration process of the traffic vehicle denoted by car1, namely, an initial lateral offset, an initial longitudinal offset, and an initial speed, are output and displayed by the computer device, as denoted by the 401 in fig. 5 b. The scene parameter that can be subjected to the generalization configuration processing means that, after the parameter value of the corresponding scene parameter is adjusted, the running result obtained by performing the scene running by adopting the adjusted scene parameter value is different, and if the result of performing the scene running by adopting the adjusted scene parameter value is the same as the result of performing the scene running before the adjustment, the scene parameter is considered to be the scene parameter that is not subjected to the generalization configuration processing, for example, the scene parameter that can be subjected to the generalization configuration processing can be acceleration or speed, etc., and the scene parameter that can not be subjected to the generalization configuration processing can be the length of a lane line or the width of a sidewalk, etc. In another implementation, the computer device may also use all scene parameters of the current configuration object as parameters that may be subjected to the generalization configuration process to achieve accurate description of the real test scene. In one embodiment, the laterally offset parameter value refers to a position where the corresponding traffic participant (e.g., the traffic vehicle named car1 above) can be offset to the left (or right) based on the current travel position, while the longitudinally offset parameter value refers to a position where the corresponding traffic participant is offset forward (or backward) based on the current travel position support, and the initial speed parameter value refers to the travel speed.

After the computer equipment obtains one or more scene parameters of the current configuration object, the operation object can respectively set a minimum value, a maximum value and a generalization interval of the corresponding scene parameters for each scene parameter, so that the computer equipment can obtain the maximum value, the minimum value and the generalization interval set by the operation object for any scene parameter of the current configuration object according to the setting of the operation object in the parameter generalization interface, and further can generate a parameter generalization rule of any scene parameter of the current configuration object according to the maximum value, the minimum value and the generalization interval set for any scene parameter of the current configuration object. As shown by a 401 label interface in fig. 5b, the parameter generalization rules corresponding to the traffic vehicle named car1 obtained by the computer device respectively include corresponding parameter generalization rules obtained by generalizing configuration processing on an initial lateral offset, an initial longitudinal offset and an initial speed, wherein the parameter generalization rules corresponding to the initial lateral offset can be values in a range of 0-0.1m, and can be values according to an interval of 0.01m when the values are obtained, the parameter generalization rules corresponding to the initial longitudinal offset can be values in a range of 0-0.2m, and can be values according to an interval of 0.01m when the values are obtained, and the parameter generalization rules corresponding to the initial speed can be values in an interval of 25-100 m/s, and can be values according to an interval of 1 m/s.

In one embodiment, the process of performing the generalization configuration processing on each scene parameter in the host vehicle named planner0 and the traffic vehicle named car2 in the template scene named base_abc_001 may be similar to the process of performing the generalization configuration processing on each scene parameter in the traffic vehicle named car1 described above, and will not be described herein. The scene parameters of the computer device for generalization configuration processing can further comprise one or more of the following: the parameters related to the corresponding simulation scene, such as acceleration, various events (such as distance, speed, collision time (Time To Collision, TTC), time, absolute position, etc.), the number of the lane to which the trigger parameters belong, etc., and the scene parameters included in the traffic participation objects in each template scene may be the same or different, which is not limited in the embodiment of the present application.

In one embodiment, when the computer device performs the generalization configuration processing on the scene parameters of the traffic participation objects included in the template scene, the computer device may also add or delete the traffic participation objects included in the template scene, and similarly, may also add or delete the scene parameters of the corresponding traffic participation objects in the template scene, so as to obtain a diversified and rich generalization simulation scene in the subsequent generalization processing stage.

After the computer equipment obtains the parameter generalization rule corresponding to each scene parameter in the template scene, a generalization parameter storage module in the computer equipment can generate corresponding storage according to the parameter generalization rule corresponding to each scene parameter to store. Wherein, as in the above-mentioned parameter generalization rule of the three scene parameters related to the traffic participation object named car1 in fig. 5b, based on the value range and the value interval of the corresponding scene parameters specified by the parameter generalization rule, 10×20×76=15200 possible values of the traffic participation object named car1 exist, that is, 15200 different generalization simulation scenes are generated by the traffic participation object support combination named car1, further, the computer device may obtain the possible value of the traffic participation object named planner0, assume 100, and obtain the possible value of the traffic participation object named car2, assume 200, then the computer device may obtain the estimated number of the generalized simulation scenes supported by the template scene by randomly combining the possible values of the traffic participation object existence, wherein the estimated number is: 15200×100×200=304×10 ⁶ And (5) generalizing the simulation scene.

After obtaining the estimated number of the generalized simulation scenes supported to be generated, the computer equipment obtains an estimated space for storing the generalized simulation scenes supported to be generated according to the estimated number, wherein the estimated space is formed by the computer equipment, and the estimated space is a space for storing the generalized simulation scenes supported to be generatedIf the estimated number of the generalized simulation scenes is 304×10 ⁶ And the storage space required for storing one generalized simulation scene is 1KB, then the total space required by the computer equipment for storing the estimated number of generalized simulation scenes is 304×10 ⁶ KB。

The estimated number of the generalized simulation scenes supported to be generated by the computer device (the estimated number may be 304×10 as described above ⁶ ) And a memory space (e.g. 304 x 10 as described above) required for storing the estimated number of generalized simulation scenes ⁶ KB), the computer device may further obtain a scene identifier (such as base_abc_001 described above) of the template scene and object identifiers (such as planner0, car1 and car2 described above) of each traffic participant object in the template scene, so that the computer device may generate generalization information data according to the scene identifier of the template scene, the object identifiers of each traffic participant object in the template scene, the parameter generalization rule, the predicted number and the predicted space of the scene parameter of each traffic participant object, and may further locally store the generalization information data. Based on the storage of the computer device for the generalization information data, the generalization capability of performing scene generalization processing by adopting a corresponding template scene can be determined based on the estimated number recorded in the generalization information data, the storage space required by the generated generalization simulation scene can be determined based on the estimated space, and then the number of the generated generalization simulation scenes based on the template scene can be determined by combining the storage capability of the device and the storage space required by the estimation, for example, when a scene exporting device (such as a terminal device) is exporting the generalization simulation scene, the template scene for executing export operation can be determined by combining the estimated space and the storage capability of the device, and/or the total number of exported generalization simulation scenes can be determined, for example, if the self-storage capability of the scene exporting device only supports 6×10 ⁶ The generalized simulation scene of KB is stored, and the generalized information data 1 is assumed to indicate that the estimated storage space of the generalized simulation scene of the template scene 1 is 304×10 ⁶ KB, the generalized information data 2 indicates that the estimated storage space of the generalized simulation scene of the template scene 2 is 4×10 ⁶ KB, then the scene exports the device baseThe self-energy storage capacity can determine that the template scene 2 is adopted to carry out scene generalization processing, or the scene derivation equipment can also adopt the template scene 1 to continuously generate the generalization simulation scene after adopting the template scene 2 to carry out generalization processing until the obtained generalization simulation scene occupies 6 x 10 completely ⁶ The memory space of KB, or the scene derivation device may also determine, based on its own memory capability, that the scene generalization process is performed using the template scene 1, but when the scene generalization process is performed using the template scene 1, the memory space corresponding to the simulation scene generated by the generalization reaches 6×10 ⁶ After KB, the process of generalization processing is stopped.

In one embodiment, since the storage of the generalization information data by the computer device is a corresponding field of the storage, and the storage space occupied by each field is less, the storage resource of the computer device can be greatly released based on the storage of the generalization information data without occupying more storage resource. In addition, when the generalization information data is stored in the database, the database can be a database in the computer equipment, so that the generalization information data is stored locally in the computer equipment, or the database can be a database of a cloud platform, so that the generalization information data is stored in the cloud platform.

In addition, based on the storage of the generalization information data by the generalization parameter storage module, the computer equipment can also support the derivation of the simulation scene, so that the mass generalization scene can be downloaded to the local area by the related objects from the cloud platform (or the computer equipment), or the generalization scene can be shared and sent to other objects. Specifically, the generalized scene derivation module may obtain a scene derivation requirement from the target device, where the scene derivation requirement is used to indicate the number of generalized simulation scenes to be derived or template scenes for performing scene derivation, and after obtaining the scene derivation requirement, may generate a derivation file according to the scene derivation requirement information, where the generated derivation file includes one or more of the following: the scene generation tool is used for performing scene derivation on the target template scene and the corresponding parameter generalization rule. After the computer equipment generates a corresponding export file based on the scene export requirement, the export file can be sent to the target equipment, so that the target equipment can generate a generalization simulation scene according to the target template scene and a corresponding parameter generalization rule by using a scene generation tool in the export file, and the required generalization simulation scene is downloaded to the local of the target equipment. It can be understood that when the scene derivation requirement is the total number of scenes of the generalized simulation scene to be derived, the target template scene is one template scene, and when the scene derivation requirement is the indication information of the template scene used for generating the generalized simulation scene, the target template scene is the template scene used for generating the generalized simulation scene in the plurality of template scenes. Even if the target equipment needs to generate a large number of generalized simulation scenes, the target equipment only comprises a template scene file, parameter generalization rules, scene generation tools and other small documents when the target equipment is exported, so that the export speed of the large number of generalized simulation scenes is greatly improved, a large amount of network bandwidth and storage space are saved, in addition, related users only need to generate specific generalized simulation scenes through the scene generation tools when the related users need to generate the specific generalized simulation scenes, and a large amount of storage space is not wasted due to the fact that the large number of generalized scene files are reserved at ordinary times, so that the storage resources of the target equipment are saved.

S404, acquiring scene operation requirements, and acquiring parameter generalization rules related to the scene operation requirements from a database, wherein the parameter generalization rules are used for determining one or more traffic participation objects to be generalized and the value ranges of corresponding scene parameters; wherein, a traffic participation object to be generalized corresponds to one or more scene parameters, and a scene parameter has a corresponding parameter generalization rule.

In one embodiment, when the capability test of the autopilot algorithm is required, the computer device can determine that the scene operation requirement exists because the autopilot algorithm needs to be carried into a host vehicle of a corresponding simulation scene to operate, thereby realizing the capability test process of the autopilot algorithm based on the operation process of the simulation scene.

In one embodiment, the flooding demand information is determined by a parameter flooding rule that is operative to be included in the flooding information data stored in the database, wherein, since one of the flooding information data stored in the database corresponds to one of the template scenes, if the one of the flooding information data stored in the database indicates that the one of the template scenes is related to the generation of the flooding scene, that is, the parameter flooding rule included in the one of the flooding information data stored in the database indicates that the parameter flooding rule corresponding to each of the scene parameters included in the one of the template scenes is included, the computer device may select a part or all of the parameter flooding rules included in the one of the flooding information data stored in the database based on all of the parameter flooding rules included in the one of the flooding information data stored in the database, and use the selected parameter flooding rule as the flooding demand information to perform the operation of the flooding simulation scene. In another implementation manner, in the process that the computer device determines the number of the predictions included in the generalization information data of the corresponding template scene in advance, the computer device may further perform numbering processing on the generalization simulation scene supported by the template scene, for example, the number of the generalization simulation scene supported by one template scene may be sequentially 1-500, and then the operation object may perform subsequent generation of the generalization simulation scene by inputting the corresponding scene number into the computer device and using the input one or more scene numbers as the generalization requirement information, so as to implement the generation process of the customized simulation scene. Or, in yet another implementation manner, the operation object may further send, to the computer device, a target value less than or equal to a value corresponding to the estimated number based on the estimated number value included in the generalization information data of the corresponding template scene, so that the computer device generates, based on the obtained target value, a generalization simulation scene equivalent to the target value based on the corresponding parameter generalization rule. It can be understood that, when the generalization information data stored in the database by the computer device is only the generalization information data generated by a certain template scene, the generalization requirement information acquired by the computer device can be used for indicating the total number of scenes of the generalization simulation scene to be generated.

In one embodiment, if the generalization information data stored in the database includes generalization information data corresponding to at least two template scenes, the operation object may directly select part or all of the generalization information data, so that the computer device may use the selected generalization information data as generalization requirement information and perform generation of a generalization simulation scene based on the selected generalization information data, and it may be understood that when the computer device performs generation of a generalization simulation scene based on the selected generalization information data, the computer device performs generation of a simulation scene on each generalization parameter rule included in the selected generalization information data, that is, when there is generalization information data corresponding to at least two template scenes in the database, the obtained generalization requirement information by the computer device indicates one or more template scenes used for performing generation of the generalization simulation scene.

S405, performing generalization processing on corresponding scene parameters according to the acquired parameter generalization rules, and generating one or more generalized simulation scenes.

The generation of the generalization simulation scene by the computer device can be performed by calling a generalization scene generation module in the computer device, and in the embodiment of the application, the description will be mainly made on the case that the generalization information data stored in the database is the generalization information data corresponding to one template scene, and when the generalization information data in the database contains the generalization information data corresponding to different template scenes, the embodiment of the application can be also seen. In a specific implementation, the computer device may first use one or more traffic participation objects included in the template scene as traffic participation objects to be subjected to generalization processing, and use scene parameters of any traffic participation object to be subjected to generalization processing included in the template scene as scene parameters of corresponding traffic participation objects in the generalization simulation scene; as shown in fig. 5a, if the traffic participation object in the template scene is specifically named as a host vehicle of planner0 and traffic vehicles of car1 and car2 respectively, then the computer device takes the host vehicle of planner0 and the traffic vehicles of car1 and car2 respectively as the traffic object to be processed in a flooding manner, and takes the scene parameters corresponding to the host vehicle of planner0 and the scene parameters corresponding to the traffic vehicles of car1 and car2 respectively as the scene parameters in the flooding simulation scene, that is, the host vehicle of planner0 and the traffic vehicles of car1 and car2 respectively are also contained in the flooding simulation scene obtained by the computer device based on the template scene, and the scene parameters of the host vehicle of planner0 and the traffic vehicles of car1 and car2 respectively are also the same as the scene parameters contained in the template scene.

In another implementation, the computer device may also take part of the traffic participation objects included in the template scene as the traffic participation objects to be flooded and take the corresponding scene parameters as the scene parameters of the flooding simulation scene, that is, if the traffic participation objects included in the template scene include a host vehicle named planner0 and traffic vehicles named car1 and car2, respectively, the computer device may select one or two (e.g., only the host vehicle or the host vehicle and the traffic vehicle named car 1) from among the host vehicles named planner0 and the traffic vehicles named car1 and car2, respectively, as the traffic participation objects to be flooded and take the corresponding scene parameters as the scene parameters of the corresponding traffic participation objects.

After the computer equipment determines the traffic participation objects and the corresponding scene parameters of the generalized simulation scene, the corresponding scene parameters can be randomly valued according to the parameter generalization rule, and the randomly valued scene parameters are randomly combined to generate one or more generalized simulation scenes. When the computer equipment carries out random value taking on the corresponding scene parameters according to the parameter generalization rule, the maximum value, the minimum value and the generalization interval contained in the parameter generalization rule corresponding to any scene parameter can be obtained first, then the random value taking can be carried out on any scene parameter according to the generalization interval within the value range specified by the maximum value and the minimum value, and finally the random value taking processing is carried out on the parameter values after the random value taking and the corresponding relation between the traffic participation objects and the scene parameters, so that the generation of one or more generalization simulation scenes is realized.

In one embodiment, the one or more generalization simulation scenes obtained by the generalization process are obtained through one or more rounds of dynamic generalization processes, that is, when the computer device generates the generalization simulation scenes, the number of the generalization simulation scenes generated based on one round of the generalization process can be determined by combining with the concurrency number supported by the generalization scene operation module. The process by which the computer device generates one or more generalized simulation scenarios is described in detail below in conjunction with FIG. 5 c. Firstly, a computer device can establish an operation task of a simulation scene after acquiring the information of the generalization requirement, then the computer device can read parameter generalization rules from a database based on the information of the generalization requirement, wherein the computer device can read the parameter generalization rules of which the number is required for executing one round of the generalization processing all at a time, or can read only the parameter generalization rules of which the number is required for executing one round of the generalization processing only, in one embodiment, if the parameter generalization rules read by the computer device are the number of complications of the scene operation which are acquired first, then the corresponding scene parameters can be subjected to the generalization processing by adopting the parameter generalization rules with the same quantity as the concurrence quantity, so that the corresponding scene parameters with the same quantity as the concurrence quantity are obtained, then the acquisition of the generalization stop condition is carried out, and whether the next round of the generalization processing is executed is determined according to the generalization stop condition. In one embodiment, when the generalization requirement information for indicating the running requirement of the scene is used to determine the total number of scenes of the generalization simulation scene to be generated, the generalization stop condition is: the total number of the generalized simulation scenes obtained after one or more rounds of generalization processing is more than or equal to the total number of the scenes of the generalized simulation scenes to be generated; and when the generalization demand information is used for indicating a template scene used for generating a generalization simulation scene, the generalization stopping condition is as follows: the total number of the generalized simulation scenes obtained after one or more rounds of generalization processing is greater than or equal to the total number of the generalized simulation scenes generated by the support of a plurality of template scenes.

That is, when the computer device determines that the total number of the generated generalized simulation scenes is equal to the required total number of the scenes based on the acquired generalized stop conditions, or when the generation of the generalized simulation scenes supported and generated for all the selected template scenes is completed, the process of reading the parameter generalization rule and performing the generalization processing according to the read parameter generalization rule may be stopped, otherwise, the computer device may send the generated generalized simulation scenes of the round to the generalized simulation scene operation module after generating the one round of the generalized simulation scenes, and concurrently operate the generated generalized simulation scenes after performing the one round of the generalized processing by the generalized simulation scene operation module, and iteratively perform the acquisition of the generalized stop conditions after the operation of the generated generalized simulation scenes is completed by performing the one round of the generalized processing, and then determine whether the operation of the one round of the generalized simulation scenes needs to be performed again until the generalized stop conditions are satisfied, and stop the generation process of the generalized simulation scenes. Therefore, when the computer equipment generates the generalized simulation scenes, the generalized scene generating module adopts a dynamic generating mode to execute, and after the generalized scene operating module operates a certain number of scenes, part of the generalized scenes are generated again for the scene operating module to operate and execute, that is, the generation of a large number of generalized simulation scenes is not completed at one time, but the generation and operation processes of a small number of generalized simulation scenes are continuously repeated, so that the problem that the computer equipment needs to store the generated generalized simulation scenes in advance, and a large amount of storage space is consumed due to the fact that the computer equipment needs to store a large number of generalized simulation scene files is solved.

S406, running the generated generalized simulation scene.

Because the computer equipment dynamically generates a large number of generalized simulation scenes in a multi-round iteration mode, the computer equipment acquires an operation result of one generalized simulation scene after the one generalized simulation scene is operated and generates operation information according to the operation result. When the computer device runs the generalized simulation scene, the automatic driving algorithm is carried into the host vehicle corresponding to the generalized simulation scene to run, and then the running result obtained based on the running of the generalized simulation scene can be used for indicating the evaluation result of the automatic driving algorithm, namely the index value of each evaluation index of the automatic driving algorithm, after the corresponding automatic driving algorithm is carried in one generalized simulation scene, and when the computer device generates the running information based on the obtained running result, the generated running information can further comprise one or two of the following: scene numbers of the corresponding generalized simulation scenes, and parameter values of each scene parameter of the corresponding generalized simulation scenes. In one embodiment, after obtaining the operation information of each generalized simulation scene, the computer device may store the operation information corresponding to each generalized simulation scene, and may perform a query of the generalized scene and a query process of testing performance of the autopilot algorithm based on the operation information later.

After the computer equipment stores the corresponding operation information of each generalized simulation scene, in order to further release the storage resources of the computer equipment, the computer equipment can delete the corresponding generalized simulation scene after storing the operation information of the corresponding generalized simulation scene, thereby further reducing the data storage pressure of the computer equipment, realizing the effective saving of the storage resources and also realizing the effective improvement of the computing capacity of the computer equipment. In one embodiment, after storing the operation information of each generalized simulation scene, the generalized scene query module may further provide a corresponding query interface for the operation user, where the query interface may be an interface as shown in fig. 5d, and then the computer device may obtain the scene query information from the query interface, find the operation information matching the scene query information from the stored operation information, and output and display the found operation information. The scene query information which can be input by the operation user in the query interface can be one or two of a target scene number and a target parameter value, the operation information of the target generalized simulation scene corresponding to the target scene number can be queried and obtained based on the target scene number, the operation information of one or more generalized simulation scenes containing the target parameter value can be queried and obtained based on the target parameter value, and the computer equipment outputs and displays the searched operation information, namely, the process of outputting and displaying the operation result of the automatic driving algorithm in the corresponding generalized simulation scene.

In one embodiment, practice shows that based on the above process of generating, storing, running and querying simulation scenes, a huge amount of generalized simulation scenes can be stored in a very small storage space, for example, 3.43 hundred million generalized simulation scenes under a conventional scheme can be stored by using less than 5KB of storage space by adopting the technical scheme, and about 774GB of storage space is occupied.

In the embodiment of the application, after the template scene is acquired, the computer equipment can obtain the parameter generalization rule of each scene parameter by carrying out generalization configuration processing on the scene parameter of each traffic participation object in the template scene, so that the generalization information data generated by the generalization parameter rule can be stored, the storage of only the template scene file and related generalization information data is realized, and when the operation of the simulation scene is required, the computer equipment can dynamically carry out scene generalization processing in real time based on the stored template scene file and the generalization information data, and the generated generalization simulation scene is immediately discarded after the operation of the generated generalization simulation scene is finished, so that the storage space of the computer equipment can be greatly saved. And based on the storage of the template file and the generalization information data by the computer equipment, a large number of generalization scene files can be generated in advance, the local storage space of a user can be greatly saved, the downloading speed of the massive generalization scene files can be greatly improved, and the generation and operation stages of the generalization simulation scene are facilitated. In addition, as the computer equipment generates a massive generalized simulation scene in a distributed concurrency and dynamic running mode, and the query method of the massive generalized test scene is provided, the storage space of the massive scene is greatly reduced, the generation efficiency of the massive scene is improved, the massive scene query and running efficiency is not influenced, the storage space cost is greatly reduced, and the generalized generation efficiency of the massive scene is improved.

Based on the description of the above embodiment of the scenario running method, the embodiment of the present invention also proposes a scenario running apparatus, which may be a computer program (including program code) running in the above computer device. The scenario execution device may be used to execute the scenario execution method as described in fig. 2 and 4, please refer to fig. 6, and the scenario execution device includes: an acquisition unit 601 and a processing unit 602.

The acquiring unit 601 is configured to acquire a scene operation requirement, and acquire a parameter generalization rule related to the scene operation requirement, where the parameter generalization rule is used to determine a value range of one or more traffic participation objects to be generalized and corresponding scene parameters; wherein, a traffic participation object to be generalized corresponds to one or more scene parameters, and a corresponding parameter generalization rule exists for one scene parameter;

the processing unit 602 is configured to perform generalization processing on corresponding scene parameters according to the obtained parameter generalization rule, and generate one or more generalized simulation scenes;

the processing unit 602 is further configured to run the generated generalized simulation scenario.

In one embodiment, the obtaining unit 601 is further configured to obtain a template scene, where the template scene includes one or more traffic participation objects;

The processing unit 602 is further configured to perform generalization configuration processing on the scene parameters of each traffic participation object, so as to obtain a parameter generalization rule corresponding to each scene parameter of each traffic participation object;

the processing unit 602 is further configured to generate generalization information data according to a parameter generalization rule corresponding to each scene parameter of each traffic participant in the template scene, and store the generalization information data in a database, so as to obtain a parameter generalization rule related to the scene operation information from the database.

In one embodiment, the processing unit 602 is specifically configured to, when performing the generalization configuration processing on the scene parameters of any traffic participant to obtain the parameter generalization rule corresponding to any scene parameter:

determining a current configuration object from one or more traffic participation objects contained in the template scene, and outputting one or more scene parameters of the current configuration object;

obtaining a maximum value, a minimum value and a generalization interval set for any scene parameter of the current configuration object, and generating a parameter generalization rule of any scene parameter of the current configuration object according to the maximum value, the minimum value and the generalization interval set for any scene parameter of the current configuration object;

The generalization interval is used for indicating a value interval when any scene parameter of the current configuration object is subjected to generalization processing.

In an embodiment, the obtaining unit 601 is further configured to obtain generalization requirement information, where the generalization requirement information is used to indicate a total number of scenes of a generalization simulation scene to be generated, or the generalization requirement information is used to indicate one or more template scenes used for generating the generalization simulation scene;

the processing unit 602 is further configured to determine that the scene operation requirement is acquired when the generalization requirement information is successfully acquired.

In one embodiment, the processing unit 602 is specifically configured to:

determining the estimated number of the generalized simulation scenes supported to be generated by adopting the parameter generalization rule according to the parameter generalization rule corresponding to each scene parameter of each traffic participation object in the template scene, and obtaining an estimated space for storing the generalized simulation scenes supported to be generated according to the estimated number;

acquiring a scene identifier of the template scene and an object identifier of each traffic participation object in the template scene;

and generating generalization information data according to the scene identification of the template scene, the object identification of each traffic participation object in the template scene, the parameter generalization rule of the scene parameter of each traffic participation object, the estimated quantity and the estimated space.

In one embodiment, the processing unit 602 is specifically configured to:

taking one or more traffic participation objects contained in the template scene as traffic participation objects to be subjected to generalization processing, and taking scene parameters of any traffic participation object to be subjected to generalization processing contained in the template scene as scene parameters of corresponding traffic participation objects in the generalization simulation scene;

and randomly taking values of the corresponding scene parameters according to the parameter generalization rules, and randomly combining the randomly-taken scene parameters to generate one or more generalization simulation scenes.

In one embodiment, the processing unit 602 is specifically configured to:

obtaining a maximum value, a minimum value and a generalization interval contained in a parameter generalization rule corresponding to any scene parameter;

and randomly taking values of any scene parameter according to the generalization interval in a value range specified by the maximum value and the minimum value.

In one embodiment, the processing unit 602 is further configured to obtain a scenario derivation requirement from the target device, and generate a derivation file according to the scenario derivation requirement information, where the derivation file includes one or more of the following: the scene generation tool is used for conducting scene derivation on the target template scene and the corresponding parameter generalization rule;

The processing unit 602 is further configured to send the export file to the target device, so that the target device generates a generalized simulation scene according to the target template scene and a corresponding parameter generalization rule by using a scene generating tool in the export file.

In one embodiment, the number of template scenes is one or more; the processing unit 602 is further configured to, when the scene derivation requirement is a total number of scenes of the generalized simulation scene to be derived, the target template scene is the one template scene;

the processing unit 602 is further configured to, when the scene derivation requirement is indication information of a template scene used for generating a generalized simulation scene, the target template scene is a template scene used for generating the generalized simulation scene from the plurality of template scenes.

In one embodiment, one or more generalization simulation scenes obtained by the generalization process are obtained by one or more rounds of dynamic generalization process; the acquiring unit 601 is further configured to acquire a concurrency number when performing scene operation;

the processing unit 602 is further configured to perform generalization processing on corresponding scene parameters by using a parameter generalization rule equal to the concurrency number to obtain a generalization simulation scene equal to the concurrency number;

The obtaining unit 601 is further configured to obtain a generalization stop condition, and determine whether to execute a next round of generalization processing according to the generalization stop condition.

In one embodiment, when the generalization requirement information for indicating the running requirement of the scene is used for determining the total number of scenes of the generalization simulation scene to be generated, the generalization stopping condition is: the total number of the generalized simulation scenes obtained after one or more rounds of generalization processing is more than or equal to the total number of the scenes of the generalized simulation scenes to be generated;

when the generalization requirement information is used for indicating a template scene used for generating a generalization simulation scene, the generalization stopping condition is as follows: the total number of the generalized simulation scenes obtained after one or more rounds of generalization processing is greater than or equal to the total number of the generalized simulation scenes generated by the support of a plurality of template scenes.

In one embodiment, the processing unit 602 is specifically configured to:

concurrently running the generalization simulation scene generated after the one-round generalization processing is executed;

and triggering and executing the step of acquiring the generalization stop condition after the operation of the generalization simulation scene generated by executing one round of generalization processing is finished.

In one embodiment, the processing unit 602 is further configured to obtain an operation result of one generalized simulation scene after the one generalized simulation scene is operated, and generate operation information according to the operation result; the operation result is used for indicating the evaluation result of the automatic driving algorithm after the corresponding automatic driving algorithm is carried in the generalized simulation scene;

The processing unit 602 is further configured to store operation information corresponding to each generalized simulation scene, and delete the corresponding generalized simulation scene after storing the operation information of the corresponding generalized simulation scene.

In one embodiment, the operation information further includes one or both of the following: scene numbers of the corresponding generalized simulation scenes and parameter values of various scene parameters of the corresponding generalized simulation scenes;

the acquiring unit 601 is further configured to acquire scene query information, where the scene query information is one or two of a target scene number and a target parameter value;

the processing unit 602 is further configured to find out operation information matching with the scene query information from the stored operation information, and output the found operation information.

In the embodiment of the present application, after the obtaining unit 601 obtains the scene operation requirement, the obtaining unit may also obtain the parameter generalization rule related to the scene operation requirement, so that the processing unit 602 may perform generalization processing on the scene parameters of the corresponding traffic participation object according to the traffic participation object to be subjected to generalization processing indicated by the obtained parameter generalization rule and the value range of the corresponding scene parameter, thereby obtaining one or more generalization simulation scenes, and after the processing unit 602 obtains one or more generalization simulation scenes, the automatic driving algorithm to be tested may be carried in the host vehicle of the generalization simulation scene, thereby implementing the test of the algorithm performance of the automatic driving algorithm. Because the processing unit 602 performs the generation of the generalization simulation scene based on the requirement in real time after the requirement of scene operation is acquired, the generation of the generalization simulation scene is not needed in advance, and further, the storage pressure of the generalization simulation scene generated in advance can be reduced, the effective release of storage resources is realized, the generation efficiency of a large number of generalization simulation scenes can be improved based on the release of the storage resources, in addition, if the traffic participation objects to be subjected to the generalization processing indicated by the parameter generalization rule adopted by the processing unit 602 during the generation of the generalization simulation scene are mostly the same traffic participation objects, the processing unit 602 can generate more similar generalization simulation scenes, and when the test of the automatic driving algorithm is performed by adopting the more similar generalization simulation scenes, the accuracy of the test of the algorithm boundary capability of the automatic driving algorithm can be improved, and when the test of the generalization simulation scene generated by the processing unit 602 based on the parameter generalization rule is larger, the automatic driving algorithm can be tested by adopting different large different generalization simulation scenes, and further, the comprehensive test of the automatic driving algorithm can be realized.

Fig. 7 is a schematic block diagram of a computer device according to an embodiment of the present invention. The computer device in the present embodiment as shown in fig. 7 may include: one or more processors 701; one or more input devices 702, one or more output devices 703 and a memory 704. The processor 701, the input device 702, the output device 703, and the memory 704 are connected by a bus 705. The memory 704 is used for storing a computer program comprising program instructions, and the processor 701 is used for executing the program instructions stored in the memory 704.

The memory 704 may include volatile memory (RAM), such as random-access memory (RAM); the memory 704 may also include a non-volatile memory (non-volatile memory), such as a flash memory (flash memory), a Solid State Drive (SSD), etc.; memory 704 may also include combinations of the above types of memory.

The processor 701 may be a central processing unit (central processing unit, CPU). The processor 701 may further comprise a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (programmable logic device, PLD), or the like. The PLD may be a field-programmable gate array (field-programmable gate array, FPGA), general-purpose array logic (generic array logic, GAL), or the like. The processor 701 may also be a combination of the above structures.

In an embodiment of the present invention, the memory 704 is configured to store a computer program, where the computer program includes program instructions, and the processor 701 is configured to execute the program instructions stored in the memory 704, to implement the steps of the corresponding method shown in fig. 2 and fig. 4.

In one embodiment, the processor 701 is configured to call the program instructions for executing:

and running the generated generalized simulation scene.

acquiring a template scene, wherein the template scene comprises one or more traffic participation objects;

Performing generalization configuration processing on scene parameters of each traffic participation object to obtain parameter generalization rules corresponding to each scene parameter of each traffic participation object;

generating generalization information data according to parameter generalization rules corresponding to each scene parameter of each traffic participation object in the template scene, and storing the generalization information data into a database so as to acquire parameter generalization rules related to the scene operation information from the database.

obtaining generalization demand information, wherein the generalization demand information is used for indicating the total number of scenes of a generalization simulation scene to be generated, or the generalization demand information is used for indicating one or more template scenes used for generating the generalization simulation scene;

and when the generalization requirement information is successfully acquired, determining to acquire the scene operation requirement.

obtaining scene export requirements from a target device, generating an export file according to the scene export requirement information, wherein the export file comprises one or more of the following: the scene generation tool is used for conducting scene derivation on the target template scene and the corresponding parameter generalization rule;

And sending the export file to the target equipment so that the target equipment can generate a generalization simulation scene according to the target template scene and the corresponding parameter generalization rule by using a scene generation tool in the export file.

In one embodiment, the number of template scenes is one or more; the processor 701 is configured to call the program instructions for executing:

when the scene export requirement is the total number of scenes of the generalized simulation scene to be exported, the target template scene is the template scene;

when the scene export requirement is the indication information of the template scene used for generating the generalization simulation scene, the target template scene is the template scene used for generating the generalization simulation scene in the plurality of template scenes.

In one embodiment, one or more generalization simulation scenes obtained by the generalization process are obtained by one or more rounds of dynamic generalization process; the processor 701 is configured to call the program instructions for executing:

acquiring the concurrency quantity of the scene operation;

performing generalization processing on corresponding scene parameters by adopting a parameter generalization rule equal to the concurrency quantity to obtain a generalization simulation scene equal to the concurrency quantity;

And acquiring a generalization stop condition, and determining whether to execute the next generalization process according to the generalization stop condition.

After one generalized simulation scene is operated, an operation result of the one generalized simulation scene is obtained, and operation information is generated according to the operation result; the operation result is used for indicating the evaluation result of the automatic driving algorithm after the corresponding automatic driving algorithm is carried in the generalized simulation scene;

storing the operation information corresponding to each generalized simulation scene, and deleting the corresponding generalized simulation scene after storing the operation information of the corresponding generalized simulation scene.

In one embodiment, the operation information further includes one or both of the following: scene numbers of the corresponding generalized simulation scenes and parameter values of various scene parameters of the corresponding generalized simulation scenes; the processor 701 is configured to call the program instructions for executing:

acquiring scene query information, wherein the scene query information is one or two of a target scene number or a target parameter value;

and searching the operation information matched with the scene inquiry information from the stored operation information, and outputting the searched operation information.

Embodiments of the present invention provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer readable storage medium and executes the computer instructions to cause the computer device to perform the method embodiments described above as shown in fig. 2 or fig. 4. The computer readable storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), or the like.

The foregoing disclosure is merely illustrative of some embodiments of the present invention and it is not to be construed as limiting the scope of the invention, as a person of ordinary skill in the art will appreciate that all or part of the above-described embodiments may be practiced with equivalent variations which fall within the scope of the invention as defined in the appended claims.

Claims

1. A method of operating a scene comprising:

and running the generated generalized simulation scene.

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

3. The method of claim 2, wherein the step of obtaining the parameter generalization rule corresponding to any one of the scene parameters according to the generalization configuration process of the scene parameters of the any one of the traffic participant objects comprises:

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

5. The method of claim 2, wherein generating the generalization information data according to the parameter generalization rule corresponding to each scene parameter of each traffic participant in the template scene comprises:

6. The method of claim 2, wherein the generalizing the corresponding scene parameters according to the obtained parameter generalizing rule and generating one or more generalized simulation scenes comprises:

7. The method of claim 6, wherein randomly taking values of the respective scene parameters according to a parameter generalization rule comprises:

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

9. The method of claim 8, wherein the number of template scenes is one or more; the method further comprises the steps of:

10. The method of claim 1, wherein the one or more generalization simulation scenarios resulting from the generalization process are obtained by one or more rounds of dynamic generalization process; the method for executing a round of generalization processing according to the acquired parameter generalization rule comprises the following steps:

acquiring the concurrency quantity of the scene operation;

11. The method of claim 10, wherein when the generalization requirement information indicating the scene running requirement is used to determine the total number of scenes of the generalization simulation scene to be generated, the generalization stop condition is: the total number of the generalized simulation scenes obtained after one or more rounds of generalization processing is more than or equal to the total number of the scenes of the generalized simulation scenes to be generated;

12. The method of claim 10, wherein the running the generated generalized simulation scene comprises:

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

14. The method of claim 13, wherein the operational information further comprises one or both of: scene numbers of the corresponding generalized simulation scenes and parameter values of various scene parameters of the corresponding generalized simulation scenes; the method further comprises the steps of:

15. A scene operation device, comprising:

16. A computer device comprising a processor, an input device, an output device and a memory, the processor, the input device, the output device and the memory being interconnected, wherein the memory is adapted to store a computer program comprising program instructions, the processor being configured to invoke the program instructions to perform the method of any of claims 1-14.

17. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program comprising program instructions which, when executed by a processor, cause the processor to perform the method of any of claims 1-14.