Continuous scene automatic arrangement method and device based on digital twinning
Through digital twin technology and OpenSCENARIO standardization documents, continuous scene automation orchestration of high-level autonomous driving cars is realized, solving the problem of insufficient scenario coverage and confidence in simulation tests, and improving testing efficiency and versatility.
Patent Information
- Application Number
- CN202410942289.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-15
- Publication Date
- 2024-11-15
AI Technical Summary
The simulation tests of existing autonomous driving cars have insufficient confidence and scenario coverage, which cannot effectively meet the testing needs of high-level autonomous driving, especially in dangerous working conditions and continuous scenarios.
Using a continuous scene automation orchestration method based on digital twins, continuous scene lists and standardized files are generated through the collection of static road elements and dynamic scene data, digital twin modeling, environmental layer matching and OpenSCENARIO standardized file format, and the automation of scenes is achieved. Orchestration.
It greatly improves scene coverage, shortens testing time, and improves testing efficiency. The generated standardized files can be used in combination with simulation software that supports OpenX series, which are versatile and suitable for high-level autonomous driving testing needs.
Smart Images

Figure CN118965694A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of simulation testing technology, and in particular to a method and device for automatic arrangement of continuous scenarios based on digital twins. Background Art
[0002] High-level autonomous vehicles have now entered the stage of large-scale testing and demonstration applications. Safety testing of high-level autonomous vehicles remains a core technical problem that needs to be solved urgently for their commercial application. Compared with traditional vehicles and Advanced Driver Assistant Systems (ADAS), safety testing of high-level autonomous vehicles faces new changes and challenges in test objects and test standards.
[0003] Virtual simulation technology is an indispensable technical means for automobile R&D, manufacturing, verification and testing. It can effectively shorten the technology and product development cycle and reduce R&D costs. With the development of automobile intelligence and networking, virtual simulation technology has a greater room for development. Virtual simulation testing is a key link in realizing the application of high-level autonomous driving. Vehicles with autonomous driving functions must undergo a large number of virtual simulation tests and real vehicle road tests before they can be commercialized. At present, the testing methods for autonomous driving L2 and below vehicles are mainly divided into road-based real testing and simulation-based virtual testing methods. L3 and above autonomous driving mainly adopts the following three testing methods: model and simulation testing, closed road scenario testing and open road testing. These three technologies provide a multi-faceted autonomous driving testing framework with different degrees of test control and different degrees of test environment fidelity.
[0004] Generally speaking, the current testing of autonomous driving vehicles is divided into simulation testing and road testing, but both have their own shortcomings. For simulation testing, the confidence of simulation testing and the coverage of simulation testing scenarios are the two major pain points of simulation testing. Since simulation testing is not a real test, it is difficult for people to guarantee its confidence. The confidence of simulation testing can only be guaranteed after a large number of tests (simulation testing and road testing). At the same time, when conducting simulation testing, most of them are carried out under safe conditions (that is, the driver drives safely), which means that the reliability of autonomous driving cannot be guaranteed under dangerous conditions, and many scenarios in autonomous driving simulation testing do not consider the problem of scenario coverage. In addition, the current simulation testing has a single test scenario and can only test one scenario in a section of road. It does not conform to the characteristics of many scenarios on real roads that often exist continuously and occur in a chain, which continuously affects the vehicle, and cannot meet the testing needs of high-level autonomous driving road games. Summary of the invention
[0005] In response to the defects in the prior art, the present invention provides an automatic arrangement method of continuous scenarios based on digital twins, which can greatly improve the scenario coverage, shorten the test time, and improve the test efficiency.
[0006] A method for automatic orchestration of continuous scenarios based on digital twins, comprising:
[0007] Select appropriate roads to collect static road elements and dynamic scene data;
[0008] Conduct digital twin modeling on the collected static road elements to obtain a digital twin model;
[0009] Deconstruct the digital twin model into multiple terrain points to form a terrain layer, and match the terrain layer with the weather and road surface layer to form an environmental layer;
[0010] Performing dynamic scene statistical analysis on dynamic scene data to obtain dynamic scene statistical results;
[0011] According to the dynamic scene statistics and environment layer, the rows are combined into a continuous scene list;
[0012] Determine the parameter values of the continuous scene list, use the OpenSCENARIO standardized file format to extract and write the necessary information of the continuous scene, and obtain a standardized file for generating continuous scenes.
[0013] Preferably, the weather pavement layer includes a weather type and a pavement type.
[0014] Preferably, the terrain layer and the weather pavement layer are matched to form an environmental layer, including:
[0015] The bias factor of the selected level is used to match the different weather types at a certain terrain point;
[0016] Random average matching is performed on different road surface types appearing at a certain terrain point.
[0017] Preferably, the bias factor is calculated as follows:
[0018] H(A1)=P(A1)
[0019] A is the weather type, A1 is the degree of the weather type, H(A1) is the bias factor, and P(A1) is the statistical probability of A1 occurring at a certain terrain point.
[0020] Preferably, roadside environment perception sensors are arranged to count the probability of occurrence of different weather types and degrees at different terrain points.
[0021] Preferably, necessary information for continuous scenes is extracted, including extracting corresponding parameters, directories, road networks, entities and storyboards according to the format of OpenSCENARIO standardized files, wherein parameters are used to add variables according to specific needs; directories are used to store the set parameters in advance in corresponding files; road networks are used to correspond the location of each scene to the roads in the road network; entities are used to set the status information and performance information of traffic participants; and storyboards are used to set environmental information.
[0022] Preferably, the necessary information of the continuous scene is compiled, including:
[0023] Create traffic participants and initialize the properties of traffic participants in each scene;
[0024] Write the actions of each traffic participant and add the actions of traffic participants to the same scene story;
[0025] Write trigger conditions to control the scene story;
[0026] Add the actions of traffic participants and scenario stories to the same storyboard;
[0027] Generate xosc files for dynamic continuous scenes.
[0028] Preferably, it includes a collection module, a modeling module, a matching module, a statistical module, a combination module and a generation module, wherein:
[0029] The acquisition module is used to select appropriate roads to collect static road elements and dynamic scene data;
[0030] The modeling module is used to perform digital twin modeling on the collected static road elements to obtain a digital twin model;
[0031] The matching module is used to deconstruct the digital twin model into multiple terrain points to form a terrain layer, and match the terrain layer with the weather and road surface layer to form an environment layer;
[0032] The statistical module is used to perform dynamic scene statistical analysis on dynamic scene data to obtain dynamic scene statistical results;
[0033] The combination module is used to combine and form a continuous scene list according to the dynamic scene statistical results and the environment layer;
[0034] The generation module is used to determine the parameter values of the continuous scene list, and adopts the format of the OpenSCENARIO standardized file to extract and write the necessary information of the continuous scene to obtain a standardized file for generating the continuous scene.
[0035] The beneficial technical effects of the present invention are as follows:
[0036] (1) The present invention constructs a simulation map through digital twin technology, fully considers the characteristics of terrain points, and reasonably combines weather, road surface, and dynamic scenes, which conforms to the arrangement rules of real road scenes and closely meets the requirements of high-level autonomous driving tests;
[0037] (2) The present invention arranges scenarios in a continuous form, which can greatly improve scenario coverage, shorten test time, and improve test efficiency;
[0038] (3) The standardized file for generating continuous scenes obtained by the present invention in accordance with the format of the OpenSCENARIO standardized file can be used in combination with simulation software supporting the OpenX series and can be called repeatedly, so it is easy to use and has good versatility. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the prior art description. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.
[0040] Figure 1 A flowchart of a method for automatic orchestration of continuous scenes based on digital twins provided in one embodiment of the present invention;
[0041] Figure 2 An overall framework diagram of a method for automatic orchestration of continuous scenes based on digital twins provided in one embodiment of the present invention;
[0042] Figure 3 A dynamic scene statistical proportion map of a certain terrain point provided by an embodiment of the present invention;
[0043] Figure 4 A distribution statistics diagram of parameter values of a continuous scene list provided by an embodiment of the present invention;
[0044] Figure 5 A distribution statistics diagram of parameter values of a continuous scene list provided by another embodiment of the present invention;
[0045] Figure 6 A distribution statistics diagram of parameter values of a continuous scene list provided by another embodiment of the present invention;
[0046] Figure 7 A structural schematic diagram of a continuous scene automatic orchestration device based on digital twin is provided for one embodiment of the present invention. DETAILED DESCRIPTION
[0047] The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.
[0048] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the present invention belongs.
[0049] In one embodiment, Figure 1-Figure 2 As shown, a method for automatic orchestration of continuous scenarios based on digital twins is provided, including:
[0050] 1. Select appropriate roads to collect static road elements and dynamic scene data;
[0051] Specifically, the data collection vehicle is deployed to the selected appropriate roads to collect static road elements, including road width, slope, curve curvature, etc. The dynamic scenes on the road are also recorded, including the possible actions of following and rear-ending vehicles on the road during the day, at night and in different weather conditions. A large amount of real-time dynamic scene data is collected to facilitate statistical analysis in subsequent steps.
[0052] 2. Perform digital twin modeling on the collected static road elements to obtain a digital twin model;
[0053] In one embodiment, a 20km section of a highway in Chongqing is selected as a digital twin template. The map has typical terrain points on highways, including straight roads, curves, tunnels, service areas, toll stations, ramps, etc. Since high-level autonomous driving vehicles need to undergo actual testing on public roads before they can be officially put into use, if the simulation test map uses a single, standard, idealized structure map, it does not meet the test requirements. Therefore, the present invention selects a real road as a template (optionally an open test demonstration road) and uses a digital twin method to accurately model it.
[0054] 3. Deconstruct the digital twin model into multiple terrain points to form a terrain layer, and match the terrain layer with the weather and road surface layer to form an environment layer;
[0055] The weather type and road type of a terrain point are affected by dynamic changes and are not fixed. For example, on a certain bridge, the weather types may be day, night, rainy, or sunny, but the data collected once at a certain terrain point is only the weather type of that time. Therefore, it is necessary to collect data multiple times at different terrain points (unordered) and put them into the digital twin model, and then get all possible situations of a terrain point from the model (ordered), which is much faster and more convenient than manually screening all possible situations of a terrain point.
[0056] The terrain layer includes the types and quantities of terrain points; the types of terrain points include: straight roads, curves, bridges, tunnels, interchanges, complex sections, congestion points, steep slopes, ramps, special road structures, winding roads, bus stations, zebra crossings, T-junctions, roundabouts, roads with poor road conditions, gas stations (gas stations, charging stations), highways, urban expressways, urban trunk roads, signs, service areas and other typical terrain points.
[0057] In one embodiment, the weather-pavement layer includes weather types and pavement types.
[0058] The weather-pavement layer does not need to be collected. The weather-pavement layer is analyzed based on the weather and pavement type conditions that may occur at a certain terrain point.
[0059] Weather types include: sunny, cloudy, rainy, snowy, windy, foggy, light (natural light, artificial light), etc. Rain is divided into light rain, moderate rain, moderate to heavy rain, heavy rain, and rainstorm; snow is divided into light snow, moderate snow, heavy snow, and snowstorm; wind is divided into light wind, moderate wind, and strong wind; fog is divided into light fog, moderate fog, and heavy fog; light (natural light, artificial light) is divided into normal natural light, strong natural light, weak natural light, normal artificial light, weak natural light, and weak or no artificial light.
[0060] Road surface types include: normal road surface, low adhesion road surface, raised road surface, sunken road surface, uneven road surface, road surface with unclear lane markings, etc.
[0061] Since different terrain points affect the changes in weather and climate, the weather-pavement layer cannot be considered alone. It is necessary to match the weather type and pavement type of the terrain points and the weather-pavement layer and merge them into an environmental layer.
[0062] In one embodiment, the terrain layer and the weather and road surface layer are matched to form an environment layer, including: matching different weather types at a certain terrain point using a bias factor of the selected level; and performing random average matching on different road surface types at a certain terrain point.
[0063] Match terrain points with weather types and road surface types, and consider the single situation that may appear at a certain terrain point, which is consistent with the weather type and road surface type. After matching the terrain points, weather types and road surface types, a complete environment layer can be obtained. The environment layer is the environmental foundation for the operation of dynamic scenes. With the environment layer, the deployment and arrangement of dynamic scenes can be considered based on the environment layer.
[0064] For weather types, based on the roadside environment perception sensors deployed at multiple points, the probability of occurrence of different weather types at different terrain points is counted to adjust the weather type of the terrain points. For example, if the probability of heavy snow in the tunnel is 0, the tunnel terrain points will not match the heavy snow in the weather type to form a scene. If the probability of lighting change at the tunnel entrance is 100%, the tunnel entrance and exit terrain points must match the lighting level in the weather type to form a scene.
[0065] In one embodiment, the bias factor is calculated as follows:
[0066] H(A1)=P(A1)
[0067] A is the weather type, A1 is the degree of the weather type, H(A1) is the bias factor, and P(A1) is the statistical probability of A1 occurring at a certain terrain point.
[0068] 4. Perform dynamic scene statistical analysis on the collected dynamic scene data to obtain dynamic scene statistical results;
[0069] According to terrain points, dynamic scene statistical analysis is performed on the massive amount of dynamic scene data collected from the selected road sections, including classification and quantitative statistical analysis. The dynamic scenes that have occurred at a certain terrain point are sorted by frequency. For scenes with a frequency of less than 1%, they can be considered to be scenes with a low probability of occurring at this terrain point.
[0070] For example, for a certain terrain point on a 20km section of a highway in Chongqing, 2676 dynamic scenes were collected and classified and counted, accounting for Figure 3 At the same time, combined with the probability statistical model of typical natural driving scenarios (all roads), all the typical scenarios that appear frequently are listed, as shown in Table 1 below for comprehensive analysis, and the statistical results of dynamic scenarios are improved to judge all the typical scenarios that may occur at a certain terrain point as comprehensively as possible.
[0071] Table 1
[0072]
[0073]
[0074] 5. According to the dynamic scene statistics and the environment layer, a continuous scene list is formed;
[0075] In one embodiment, a 20 km section of a highway in Chongqing is selected, and a list of dynamic scenes that may occur at each terrain point as shown in Table 2 is obtained by combining the terrain points and dynamic scene statistics of the section.
[0076] Table 2
[0077]
[0078]
[0079]
[0080] According to the dynamic scene list corresponding to different terrain points, and then according to the environmental layer formed by the matching of the terrain layer formed by different terrain points and the weather and road surface layer, the dynamic scene list and the environmental layer are combined into a continuous scene list.
[0081] 6. Determine the parameter values of the continuous scene list, use the OpenSCENARIO standardized file format to extract and write the necessary information of the continuous scene, and obtain a standardized file for generating continuous scenes.
[0082] Further determine the parameter values of the continuous scene list. Specifically, a specific dynamic scene for the test can be built, and the key parameters of different dynamic scenes at different terrain points can be classified and counted using existing technical means to obtain parameter distribution, such as Figure 4-Figure 6 As shown, the parameter values of the continuous scene list are confirmed.
[0083] At the same time, refer to the value list of the same type of dynamic scene to improve the parameter value table of the continuous scene list. You can get the dynamic scenes corresponding to all terrain points and all possible value lists. During the test, you only need to randomly combine the dynamic scenes and their values for each terrain point (because the selected types and values are all reasonable results after screening and analysis) to combine them into a continuous scene.
[0084] In one embodiment, necessary information for continuous scenes is extracted, including extracting corresponding parameters, directories, road networks, entities and storyboards according to the format of OpenSCENARIO standardized files, wherein parameters are used to add variables according to specific needs; directories are used to store the set parameters in advance in corresponding files; road networks are used to correspond the location of each scene to the roads in the road network; entities are used to set the status information and performance information of traffic participants; and storyboards are used to set environmental information.
[0085] The goal of OpenSCENARIO is to become a specialized language in a specific field, which can appropriately express abstract and specific scenarios. Its main advantages are as follows: readability, composability, portability, strong reusability and extensibility. In addition, the OpenSCENARIO language allows the basic language mechanism and domain model to be easily extended to the quality of new requirements, including new traffic participants, models, attributes, composition and constraints of participants.
[0086] The main contents of the OpenSCENARIO standardization file are divided into parameters, catalog, road network, entities, and storyboard. The main information extracted is as follows:
[0087] (1) Parameters: Add variables according to specific needs, which can facilitate the subsequent scene triggering or stopping conditions;
[0088]
[0089] (2) Catalog: The set parameters can be stored in the corresponding files in advance, and can be called repeatedly in multiple files, which is convenient to use. For example, storing trajectory or route information can enable a traffic participant to drive according to the trajectory without specifying its specific route in the scene;
[0090]
[0091]
[0092] (3) Road network: The location of the scene is written based on the road network information in the road network file. The location of each scene needs to correspond to the roads in the road network.
[0093]
[0094] (4) Entities: Set the status information and performance information of traffic participants, etc. Each object in the scene is set here;
[0095]
[0096]
[0097]
[0098] (5) Storyboard: Sets environmental information. It can set the status information of the entire environment, and can also set changes to the environment at a specific location.
[0099] A scene storyboard contains at least one story. A story contains the following basic elements in a hierarchical structure:
[0100] Story: A high-level structure containing different aspects and at least one Act;
[0101] Act: defines a conditional behavior group, which contains at least one ManeuverGroup;
[0102] ManeuverGroup: Action group, answers who is doing what, connects Actor and Maneuver;
[0103] Maneuver: action, event group, contains at least one Event;
[0104] Event: an event, a conditional trigger action, including a start trigger and at least one Action;
[0105] Action: Action is used to create or modify dynamic elements of the scene.
[0106]
[0107]
[0108] In one embodiment, writing necessary information for continuous scenes includes:
[0109] (1) Create traffic participants and initialize the attributes of traffic participants in each scene, including the location, speed, etc. of traffic participants.
[0110] (2) Write the actions of each traffic participant and add them to the same scene story. The main actions are divided into longitudinal action (LongitudinalAction), lateral action (LateralAction), controller action (ControllerAction), teleport action (TeleportAction) and routing action (RoutingAction).
[0111] Among them, the longitudinal action (LongitudinalAction) and the lateral action (LateralAction) are mainly to operate the specific actions of the traffic participants according to a certain trigger condition;
[0112] The controller action (ControllerAction) can set the throttle, brake and other values of the vehicles of the traffic participants according to specific needs;
[0113] TeleportAction is mainly used at the beginning of the scene to generate specific traffic participants at specific locations in the road network. All added traffic participants need to be generated at specific locations, and each traffic participant corresponds to a teleport action.
[0114] RoutingAction sets a path that traffic participants can follow. After using the routing action, they can drive along the specific route without having to specify their horizontal and vertical actions.
[0115] (3) Write trigger conditions to control scene stories; write trigger conditions (trigger) for each scene to control the continuous triggering of the scenes.
[0116] (4) Add the actions of traffic participants and scene stories to the same storyboard; finally, import the actions of traffic participants and various scene stories into the same storyboard.
[0117] (5) Generate xosc files for dynamic continuous scenes.
[0118] In one embodiment, Figure 7 As shown, a continuous scene automatic arrangement device based on digital twin is provided, which is characterized by comprising an acquisition module, a modeling module, a matching module, a statistical module, a combination module and a generation module, wherein:
[0119] The acquisition module is used to select appropriate roads to collect static road elements and dynamic scene data;
[0120] The modeling module is used to perform digital twin modeling on the collected static road elements to obtain a digital twin model;
[0121] The matching module is used to deconstruct the digital twin model into multiple terrain points to form a terrain layer, and match the terrain layer with the weather and road surface layer to form an environment layer;
[0122] The statistical module is used to perform dynamic scene statistical analysis on dynamic scene data to obtain dynamic scene statistical results;
[0123] The combination module is used to combine and form a continuous scene list according to the dynamic scene statistical results and the environment layer;
[0124] The generation module is used to determine the parameter values of the continuous scene list, and adopts the format of the OpenSCENARIO standardized file to extract and write the necessary information of the continuous scene to obtain a standardized file for generating the continuous scene.
[0125] The present invention performs digital twin modeling based on real maps as the operating basis of continuous scenes. It specifically analyzes all terrain points, all weather and road surfaces, and all possible dynamic scenes of a certain section of road, matches the actual road and weather factors, and road surface factors, and ensures the authenticity and comprehensiveness of the scene. In addition, the present invention compares the format of the OpenSCENARIO standardized file to obtain a standardized file for generating continuous scenes, which can be used in combination with simulation software that supports the OpenX series and has good versatility. The present invention realizes the automatic arrangement of continuous scenes, which can reduce the cost of conversion between different technologies such as data conversion, scene classification, and scene standardized file generation, and can effectively improve work efficiency, reduce the manpower, time and cost required for conversion, and achieve the purpose of improving the work efficiency of L3 and above high-level autonomous driving simulation tests.
[0126] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.
Claims
1. A method for automated orchestration of continuous scenarios based on digital twins, It is characterized in that include: Select appropriate roads to collect static road elements and dynamic scene data; Performing digital twin modeling on the collected static road elements to obtain a digital twin model; Deconstructing the digital twin model into a plurality of terrain points to form a terrain layer, and matching the terrain layer with a weather and road surface layer to form an environment layer; Performing dynamic scene statistical analysis on the dynamic scene data to obtain dynamic scene statistical results; Combining the dynamic scene statistics result and the environment layer to form a continuous scene list; The parameter values of the continuous scene list are determined, and the necessary information of the continuous scenes is extracted and written using the format of the OpenSCENARIO standardized file to obtain a standardized file for generating continuous scenes.
2. A method for automatic arrangement of continuous scenes based on digital twins as claimed in claim 1, It is characterized in that The weather-pavement layer includes weather types and pavement types.
3. A method for automatic arrangement of continuous scenes based on digital twins as claimed in claim 2, It is characterized in that The matching of the terrain layer and the weather and road surface layer to form an environment layer includes: The bias factor of the selected level is used to match the different weather types at a certain terrain point; Random average matching is performed on different road surface types appearing at a certain terrain point.
4. A method for automatic arrangement of continuous scenes based on digital twins as claimed in claim 3, It is characterized in that The bias factor is calculated as follows: H(A1)=P(A1) A is the weather type, A1 is the degree of the weather type, H(A1) is the bias factor, and P(A1) is the statistical probability of A1 occurring at a certain terrain point.
5. A method for automatic arrangement of continuous scenes based on digital twins as claimed in claim 4, It is characterized in that Roadside environmental perception sensors are deployed to count the probability of occurrence of different weather types and degrees at different terrain points.
6. The method for automatic arrangement of continuous scenes based on digital twins as claimed in claim 1, It is characterized in that The extraction of necessary information for continuous scenes includes extracting corresponding parameters, directories, road networks, entities and storyboards according to the format of OpenSCENARIO standardized files, wherein parameters are used to add variables according to specific needs; directories are used to store the set parameters in advance in corresponding files; road networks are used to correspond the location of each scene to the roads in the road network; entities are used to set the status information and performance information of traffic participants; and storyboards are used to set environmental information.
7. A method for automatic arrangement of continuous scenes based on digital twins as claimed in claim 6, It is characterized in that The necessary information for continuous scene editing includes: Create traffic participants and initialize the properties of traffic participants in each scene; Write the actions of each traffic participant and add the actions of traffic participants to the same scene story; Write trigger conditions to control the scene story; Add the actions of traffic participants and scenario stories to the same storyboard; Generate xosc files for dynamic continuous scenes.
8. A continuous scene automatic arrangement device based on digital twins, It is characterized in that It includes acquisition module, modeling module, matching module, statistical module, combination module and generation module, among which: The acquisition module is used to select appropriate roads to collect static road elements and dynamic scene data; The modeling module is used to perform digital twin modeling on the collected static road elements to obtain a digital twin model; The matching module is used to deconstruct the digital twin model into multiple terrain points to form a terrain layer, and match the terrain layer with the weather and road surface layer to form an environment layer; The statistical module is used to perform dynamic scene statistical analysis on the dynamic scene data to obtain dynamic scene statistical results; The combination module is used to combine and form a continuous scene list according to the dynamic scene statistical results and the environment layer; The generation module is used to determine the parameter values of the continuous scene list, and adopt the format of the OpenSCENARIO standardized file to extract and write the necessary information of the continuous scene to obtain a standardized file for generating the continuous scene.
Citation Information
Cited By
Automatic driving situation awareness and decision-making method and system
CN120942374A
An autonomous driving context awareness and decision method and system
CN120942374B
