True value construction method and system fusing high-precision map and Internet of Vehicles technology

Abstract

The invention relates to a truth value construction method fusing a high-precision map and an Internet of Vehicles technology. The method comprises the following steps: step 1, acquiring high-precision map information and Internet of Vehicles communication information; 2, aligning the dynamic object data of the Internet of Vehicles communication information with a high-precision map coordinate system by adopting an instant positioning mapping technology, compensating the delay of the Internet of Vehicles information, and ensuring that the data time sequence is consistent, so as to complete data preprocessing; step 3, generating a lane-level drivable area and traffic rule constraints based on the high-precision map information data, increasing real-time traffic participant tracks and events provided by the Internet of Vehicles, embedding traffic signal lamp states and priority rules, and completing scene construction; and step 4, generating truth value data from the preprocessed data through the constructed scene model so as to support multi-dimensional performance evaluation of the automatic driving algorithm. According to the technical scheme provided by the invention, the communication reliability and security are improved.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a method and system for constructing a truth value that integrates high-precision maps and vehicle networking technologies. Background Art

[0002] Traditional autonomous driving truth-based systems rely on single sensors (such as lidar and cameras) to generate ground truth data. This system suffers from several drawbacks: limited coverage (e.g., weather and lighting conditions affect sensor performance); an inability to obtain real-time, global information about dynamic traffic participants (e.g., traffic light status and remote vehicle intentions); and a lack of deep integration of high-precision road semantic information (e.g., lane topology and traffic regulations). While vehicle-to-everything (V2X) technology (which connects vehicles to their surroundings, including vehicles, people, roads, and the cloud) can provide real-time traffic information, it lacks deep integration with high-precision maps. This results in insufficient modeling capabilities for complex scenarios, impacting the reliability and safety of autonomous vehicles. Summary of the Invention

[0003] In view of the above problems, the present invention provides a truth-value construction method and system that integrates high-precision maps and vehicle networking technologies to solve technical problems such as the existing technology relying on a single sensor to generate truth-value data, resulting in limited coverage, inability to obtain global information of dynamic traffic participants in real time, and lack of deep fusion of high-precision road semantic information.

[0004] The present invention provides a truth value construction method that integrates high-precision maps and vehicle networking technologies, which is applied to autonomous driving vehicles. The method includes: step 1, obtaining high-precision map information and vehicle networking communication information; step 2, using real-time positioning and mapping technology to align the dynamic object data of the vehicle networking communication information with the high-precision map coordinate system, compensate for the delay of the vehicle networking information, ensure data timing consistency, and thus complete data preprocessing; step 3, generating lane-level drivable areas and traffic rule constraints based on the high-precision map information data, adding real-time traffic participant trajectories and events provided by the vehicle networking, embedding traffic light status and priority rules, and completing scenario modeling; step 4, generating truth value data through the constructed scenario model using the preprocessed data to support multi-dimensional performance evaluation of the autonomous driving algorithm.

[0005] Furthermore, the method for obtaining high-precision map information data includes: step 111, performing topological structure analysis on the high-precision map through a curve fitting algorithm to extract vector data of lane lines and traffic signs; step 112, simplifying and smoothing the lane boundary lines or center lines to reduce the amount of calculation and improve the robustness of subsequent topological relationships; step 113, performing equidistant sampling and averaging on the lane boundary lines and extracting the axis in the Voronoi diagram to generate lane center lines; step 114, determining the connection position based on the lane connection relationship marked in the map, and dividing the continuous lane center lines into discrete lane segments at the key lane links.

[0006] Furthermore, the method for obtaining Internet of Vehicles communication information includes: step 121, decoding basic safety messages, traffic light phase and timing messages, and map messages; step 122, performing core field extraction, semantic verification, and coordinate system conversion on the decrypted basic safety messages to obtain the real-time position, speed, and heading of the vehicle; step 123, obtaining the real-time status of the traffic light through the decrypted traffic light phase and timing messages, and obtaining the high-precision map topology static data through the decrypted map messages.

[0007] Furthermore, the real-time status of the traffic light includes lane group number, light color status, timing information, intersection number, and phase state sequence.

[0008] Furthermore, the method of aligning the dynamic object data of the Internet of Vehicles communication information with the high-precision map coordinate system by using the real-time positioning and mapping technology includes: step 211, controlling the clock error of the Internet of Vehicles terminal, the vehicle-mounted sensor and the high-precision map server within a preset threshold through the second pulse signal or the precision clock synchronization standard protocol of the network measurement and control system; step 212, fusing the Internet of Vehicles data through the real-time positioning and mapping technology framework; step 213, associating the high-precision map with the dynamic object of the Internet of Vehicles to complete the alignment of the dynamic object data of the Internet of Vehicles communication information with the high-precision map coordinate system.

[0009] Furthermore, the pulse-per-second signal is a pulse-per-second signal of a global satellite navigation system, the Internet of Vehicles terminal includes an on-board unit and a road test unit; and the on-board sensor includes a laser radar and an inertial measurement unit.

[0010] Furthermore, the preset threshold is ±1ms.

[0011] Furthermore, the method for compensating for the delay of Internet of Vehicles information and ensuring data timing consistency includes: step 221, predicting the current and future states of the communication object based on the Internet of Vehicles information and the vehicle kinematic model; step 222, performing multi-source information fusion on the Internet of Vehicles information and the local perception information obtained by the on-board sensors; step 223, using 5G NR V2X technology to reduce the air interface transmission delay; step 224, using multiple communication technologies, and when one link is delayed too much or interrupted, switching to another link to improve the reliability and timeliness of the information.

[0012] Furthermore, step 3 includes: step 31, using a 3D target detection algorithm or a virtual private network to project the front view into a top-view coordinate system, decoding and outputting the lane instance ID and the travel direction vector, and simplifying the lane polygon through an iterative adaptive point algorithm; step 32, adding the signal light status, road construction area and traffic control information provided by the Internet of Vehicles; step 33, embedding priority rules and performing priority sorting to complete the scene construction.

[0013] The present invention also provides a system for truth value construction that integrates high-precision maps and vehicle networking technologies. The system includes: an information processing module, which is used to obtain high-precision map information and vehicle networking communication information, and adopts real-time positioning and mapping technology to align the dynamic object data of the vehicle networking communication information with the high-precision map coordinate system, compensate for the delay of the vehicle networking information, ensure the consistency of data timing, and thus complete data preprocessing; a scene modeling module, which is connected to the information processing module, and is used to generate lane-level drivable areas and traffic rule constraints based on high-precision map information data, add real-time traffic participant trajectories and events provided by the vehicle networking, embed traffic light status and priority rules, and complete scene modeling; a truth value generation module, which is respectively connected to the information processing module and the scene modeling module, and is used to generate truth value data through the constructed scene model of the preprocessed data to support multi-dimensional performance evaluation of the autonomous driving algorithm.

[0014] The present invention provides a method and system for constructing truth values ​​by integrating high-precision maps and vehicle networking technologies. This technical solution constructs system truth values ​​through multi-source data fusion, spatiotemporal synchronization, delay compensation, etc., thereby improving the verification accuracy and scenario coverage capabilities of autonomous driving algorithms, and solving the problems of low communication reliability and poor security in existing technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The present invention provides a flowchart of a true value construction method integrating high-precision maps and vehicle networking technology; Figure 2 Flowchart of the method for obtaining high-precision map information data provided by the present invention; Figure 3 A flow chart of the method for obtaining vehicle network communication information provided by the present invention; Figure 4 A flow chart of the method provided by the present invention for aligning dynamic object data of Internet of Vehicles communication information with a high-precision map coordinate system; Figure 5 This is a flow chart of the method for compensating for delay in Internet of Vehicles information provided by the present invention. DETAILED DESCRIPTION

[0016] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0017] Example 1: The present invention provides a method and system for constructing a truth value by integrating high-precision maps and vehicle networking technology. The system includes an information processing module, a scene modeling module, and a truth value generation module. Figure 1 As shown, the method includes the following steps.

[0018] Step 1: Obtain high-precision map information and vehicle network communication information; The Internet of Vehicles includes four communication modes: vehicle-to-vehicle communication (V2V), where communication information includes speed, location, and other information for collision warning and collaborative driving; vehicle-to-infrastructure interaction (V2I), where communication information includes signal light status and road construction; vehicle-to-pedestrian / cyclist communication (V2P), where communication information includes location information; and vehicle-to-network / cloud data interaction (V2N), where communication information includes remote diagnostic information, real-time navigation information, and traffic management information.

[0019] Step 2: Using real-time positioning and mapping technology, the dynamic object data of the IoV communication information is aligned with the HD map coordinate system, compensating for the delay of the IoV information and ensuring the consistency of the data timing, thus completing data preprocessing. Step 3: Generate lane-level drivable areas and traffic rule constraints based on HD map information data, add real-time traffic participant trajectories and events provided by the Internet of Vehicles, embed traffic light status and priority rules, and complete scenario modeling; Step 4: Generate true value data from the preprocessed data through the constructed scenario model to support multi-dimensional performance evaluation of the autonomous driving algorithm.

[0020] The present invention provides a method and system for constructing truth values ​​by integrating high-precision maps and vehicle networking technologies. This technical solution constructs system truth values ​​through multi-source data fusion, spatiotemporal synchronization, delay compensation, etc., thereby improving the verification accuracy and scenario coverage capabilities of autonomous driving algorithms, and solving the problems of low communication reliability and poor security in existing technologies.

[0021] Example 2: The present invention provides a method and system for constructing a truth value by integrating high-precision maps and vehicle networking technology. The system includes an information processing module, a scene modeling module, and a truth value generation module. Figure 1 As shown, the method includes the following steps.

[0022] Step 1: Obtain high-precision map information and vehicle network communication information; like Figure 2 As shown, the method for obtaining high-precision map information data includes the following steps.

[0023] Step 111: Perform topological structure analysis on the high-precision map using a curve fitting algorithm to extract vector data of lane lines and traffic signs. The original collected lane data is parsed using algorithms (such as Douglas-Peucker algorithm and spline curve fitting).

[0024] Step 112: simplify and smooth the lane boundary lines or center lines to reduce the amount of calculation and improve the robustness of subsequent topological relationships; Step 113 , performing equidistant sampling and averaging on the lane boundary lines and extracting the axis in the Voronoi diagram to generate the lane centerline; In step 114 , based on the lane connection relationship marked in the map and the determined connection position, the continuous lane centerline is segmented into discrete lane segments at the key lane link.

[0025] like Figure 3 As shown, the method for obtaining Internet of Vehicles communication information includes: Step 121, decoding the basic safety message, traffic light phase and timing message, and map message; The real-time status of traffic lights includes lane group number, light color status, timing information, intersection number, and phase state sequence.

[0026] Step 122: extract core fields, perform semantic verification, and perform coordinate system conversion on the decrypted basic security message to obtain the vehicle's real-time position, speed, and heading. Step 123: Obtain the real-time status of the traffic light through the decrypted traffic light phase and timing message, and obtain the high-precision map topology static data through the decrypted map message.

[0027] Step 2: Using real-time positioning and mapping technology, the dynamic object data of the IoV communication information is aligned with the HD map coordinate system, compensating for the delay of the IoV information and ensuring the consistency of the data timing, thus completing data preprocessing. Step 3: Generate lane-level drivable areas and traffic rule constraints based on HD map information data, add real-time traffic participant trajectories and events provided by the Internet of Vehicles, embed traffic light status and priority rules, and complete scenario modeling; Step 4: Generate true value data from the preprocessed data through the constructed scenario model to support multi-dimensional performance evaluation of the autonomous driving algorithm.

[0028] The present invention provides a method and system for constructing truth values ​​by integrating high-precision maps and vehicle networking technologies. This technical solution constructs system truth values ​​through multi-source data fusion, spatiotemporal synchronization, delay compensation, etc., thereby improving the verification accuracy and scenario coverage capabilities of autonomous driving algorithms, and solving the problems of low communication reliability and poor security in existing technologies.

[0029] Example 3: The present invention provides a method and system for constructing a truth value by integrating high-precision maps and vehicle networking technology. The system includes an information processing module, a scene modeling module, and a truth value generation module. Figure 1 As shown, the method includes the following steps.

[0030] Step 1: Obtain high-precision map information and vehicle network communication information; Step 2: Using real-time positioning and mapping technology, the dynamic object data of the IoV communication information is aligned with the HD map coordinate system, compensating for the delay of the IoV information and ensuring the consistency of the data timing, thus completing data preprocessing. SLAM (Simultaneous Localization and Mapping) technology is used to align V2X dynamic data with the HD map coordinate system. Through the PPS (Pulse Per Second) signal of the GNSS (Global Navigation Satellite System) or the IEEE 1588 PTP protocol (a standard protocol for precision clock synchronization of network measurement and control systems), the clock error between the V2X terminal (OBU / RSU), on-board sensors (LiDAR / IMU), and HD map server is controlled within ±1ms. V2X data is fused through the SLAM framework to associate HD maps with V2X dynamic objects, as shown in Table 1.

[0031]

[0032] Table 1 HD map and V2X dynamic object association table like Figure 4 As shown, the method of aligning the dynamic object data of vehicle network communication information with the high-precision map coordinate system using real-time positioning and mapping technology includes the following steps.

[0033] Step 211: Control the clock errors of the IoV terminals, vehicle sensors, and high-precision map servers within a preset threshold using pulse-per-second signals or a standard protocol for precise clock synchronization of a network measurement and control system. The pulse-per-second signal is a pulse-per-second signal from a global satellite navigation system. The Internet of Vehicles terminal includes an onboard unit and a road test unit. The onboard sensors include a laser radar and an inertial measurement unit. The preset threshold is ±1ms.

[0034] Step 212: Integrate the Internet of Vehicles data through the real-time positioning and mapping technology framework; Step 213: Associating the high-precision map with the dynamic object of the Internet of Vehicles, and completing the alignment of the dynamic object data of the Internet of Vehicles communication information with the coordinate system of the high-precision map.

[0035] like Figure 5 As shown, V2X dynamic data has a communication delay (50-100ms). To maintain synchronization between the two time dimensions, delay compensation is required. The method for compensating for the delay of Internet of Vehicles information and ensuring data timing consistency includes the following steps.

[0036] Step 221 , predicting the current and future states of the communication target based on the Internet of Vehicles information and the vehicle kinematics model; Step 222: Perform multi-source information fusion on the Internet of Vehicles information and the local perception information acquired by the vehicle-mounted sensors; Due to differences in data formats and semantic representation between connected vehicle information and local perception information, heterogeneous multi-source data fusion is necessary. V2X information should not be used in isolation. It should be fused with local perception information acquired by onboard sensors (cameras, radar, lidar, IMU, GPS). The perception latency of local sensors (especially radar and lidar) is typically much lower than the latency of V2X communication (milliseconds vs. tens to hundreds of milliseconds). After fusion, even with delayed V2X information, local sensors can provide a more up-to-date snapshot of the environment.

[0037] Step 223: Use 5G NR V2X technology to reduce air interface transmission delay; Measures are taken at the protocol stack level to reduce latency or prioritize critical information. 5G NR V2X (5GNR-V2X, New Radio Vehicle to Everything, is the application of 5G technology in the connected vehicle field, designed to support efficient communication between vehicles and their surroundings, including other vehicles, infrastructure, pedestrians, etc.) features uRLLC (Ultra-Reliable Low Latency Communications, a communication technology designed to provide high-reliability and low-latency communication services) and new-generation standards such as 802.11bd are used to significantly reduce air interface transmission latency through physical and MAC layer design.

[0038] Step 224: Use multiple communication technologies and switch to another link when one link is delayed too much or interrupted, so as to improve the reliability and timeliness of information.

[0039] In dense urban areas or tunnels, C-V2X PC5 direct communication can lead to increased packet loss (>10%) due to obstruction or interference. Therefore, multiple communication technologies must be employed, along with redundant transmission or data interpolation mechanisms. Utilizing multiple communication technologies simultaneously (such as PC5 direct communication combined with Uu cellular networks) allows for rapid switching to another link when latency is excessive or interrupted on one link, improving information reliability and timeliness. Cellular links (particularly 5G) can sometimes provide more stable, low-latency connections than direct links (depending on deployment and load).

[0040] Step 3: Generate lane-level drivable areas and traffic rule constraints based on HD map information data, add real-time traffic participant trajectories and events provided by the Internet of Vehicles, embed traffic light status and priority rules, and complete the scenario construction; Scenario modeling consists of three layers: a static layer, which generates lane-level drivable areas and traffic rules based on HD maps; a dynamic layer, which overlays real-time traffic participant trajectories and events (such as construction zones and emergency braking) provided by V2X; and a logical layer, which embeds traffic light status and priority rules (such as right-of-way allocation for ramp merging). The steps for building a scenario are as follows.

[0041] Step 31: Use a 3D object detection algorithm or a virtual private network to project the front view into a top-view coordinate system, decode and output the lane instance ID and travel direction vector, and simplify the lane polygon using an iterative adaptive point algorithm. The 3D object detection algorithm, Lift-Splat-Shoot (LSS), is a bottom-up algorithm for constructing Bird's Eye View (BEV) features. LSS generates a pseudo-view cone point cloud by back-projecting image features into 3D space. The EfficientNet algorithm extracts depth and image features from the cloud points and estimates the depth information. Finally, the point cloud features are converted to BEV space for feature fusion and subsequent semantic segmentation. The iterative adaptive point algorithm, also known as the Douglas–Peucker algorithm, approximates a curve as a series of points, reducing the number of points. Its advantage is translation and rotation invariance. Given a curve and a threshold, the sampling result is fixed.

[0042] Step 32: Add traffic light status, road construction zone, and traffic control information provided by the Internet of Vehicles. The traffic light status provided by the vehicle network is a dynamic data that changes in real time, and the traffic light decision rules mentioned in the next step are a series of decision rules formulated based on the traffic light status.

[0043] Step 33: embed traffic light decision rules and priority rules to perform priority sorting, thereby completing the scenario construction.

[0044] The priority rules are shown in Table 2.

[0045]

[0046] Table 2 Priority Rules Step 4: Generate true value data from the preprocessed data through the constructed scenario model to support multi-dimensional performance evaluation of the autonomous driving algorithm.

[0047] Generates 4D ground truth data (3D space + time series) with timestamps, supporting multi-dimensional performance evaluation of autonomous driving algorithms.

[0048] The present invention provides a method and system for constructing truth values ​​by integrating high-precision maps and vehicle networking technologies. This technical solution constructs system truth values ​​through multi-source data fusion, spatiotemporal synchronization, delay compensation, etc., thereby improving the verification accuracy and scenario coverage capabilities of autonomous driving algorithms, and solving the problems of low communication reliability and poor security in existing technologies.

[0049] In summary, the present invention provides a method and system for constructing a truth value that integrates high-precision maps and vehicle-to-everything (V2X) technologies. This technical solution builds a Level 3 autonomous driving truth value system based on the integration of high-precision maps and vehicle-to-everything (V2X) technologies. The coverage of truth data is improved, supporting verification in multiple scenarios such as urban roads, highways, and tunnels; the prediction error of dynamic target trajectories is reduced to within 0.1m; and it supports precise testing of Level 3 autonomous driving systems in ODD (operational design domain) boundary conditions (such as traffic light failure scenarios), which can be used to improve the verification accuracy and scenario coverage capabilities of autonomous driving algorithms.

[0050] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A truth value construction method integrating high-precision maps and vehicle networking technology, applied to autonomous driving vehicles, characterized by: The method comprises: Step 1: Obtain high-precision map information and vehicle network communication information; Step 2: Using real-time positioning and mapping technology, the dynamic object data of the IoV communication information is aligned with the HD map coordinate system, compensating for the delay of the IoV information and ensuring the consistency of the data timing, thus completing data preprocessing. Step 3: Generate lane-level drivable areas and traffic rule constraints based on HD map information data, add real-time traffic participant trajectories and events provided by the Internet of Vehicles, embed traffic light status and priority rules, and complete the scenario construction; Step 4: Generate true value data from the preprocessed data through the constructed scenario model to support multi-dimensional performance evaluation of the autonomous driving algorithm.

2. The method for constructing a true value integrating high-precision maps and Internet of Vehicles technology according to claim 1, characterized in that: The method for obtaining high-precision map information data includes: Step 111: Perform topological structure analysis on the high-precision map using a curve fitting algorithm to extract vector data of lane lines and traffic signs. Step 112: simplify and smooth the lane boundary lines or center lines to reduce the amount of calculation and improve the robustness of subsequent topological relationships; Step 113 , performing equidistant sampling and averaging on the lane boundary lines and extracting the axis in the Voronoi diagram to generate the lane centerline; In step 114 , based on the lane connection relationship marked in the map and the determined connection position, the continuous lane centerline is segmented into discrete lane segments at the key lane link.

3. The method for constructing a true value integrating high-precision maps and Internet of Vehicles technology according to claim 1, characterized in that: The method for obtaining Internet of Vehicles communication information includes: Step 121, decoding the basic safety message, traffic light phase and timing message, and map message; Step 122: extract core fields, perform semantic verification, and perform coordinate system conversion on the decrypted basic security message to obtain the vehicle's real-time position, speed, and heading. Step 123: Obtain the real-time status of the traffic light through the decrypted traffic light phase and timing message, and obtain the high-precision map topology static data through the decrypted map message.

4. The method for constructing a true value integrating high-precision maps and Internet of Vehicles technology according to claim 3 is characterized in that: The real-time status of traffic lights includes lane group number, light color status, timing information, intersection number, and phase state sequence.

5. The method for constructing a true value integrating high-precision maps and Internet of Vehicles technology according to claim 1 is characterized in that: The method for aligning dynamic object data of Internet of Vehicles communication information with a high-precision map coordinate system using real-time positioning and mapping technology includes: Step 211: Control the clock errors of the IoV terminals, vehicle sensors, and high-precision map servers within a preset threshold using pulse-per-second signals or a standard protocol for precise clock synchronization of a network measurement and control system. Step 212: Integrate the Internet of Vehicles data through the real-time positioning and mapping technology framework; Step 213: Associating the high-precision map with the dynamic object of the Internet of Vehicles, and completing the alignment of the dynamic object data of the Internet of Vehicles communication information with the coordinate system of the high-precision map.

6. The method for constructing a true value integrating high-precision maps and Internet of Vehicles technology according to claim 5 is characterized in that: The second pulse signal is a second pulse of the global satellite navigation system, the Internet of Vehicles terminal includes a vehicle-mounted unit and a road test unit; the vehicle-mounted sensor includes a laser radar and an inertial measurement unit.

7. The method for constructing a true value integrating high-precision maps and Internet of Vehicles technology according to claim 5 is characterized in that: The preset threshold is ±1ms.

8. The method for constructing a true value integrating high-precision maps and Internet of Vehicles technology according to claim 1, characterized in that: The method for compensating for the delay of Internet of Vehicles information and ensuring data timing consistency includes: Step 221 , predicting the current and future states of the communication target based on the Internet of Vehicles information and the vehicle kinematics model; Step 222: Perform multi-source information fusion on the Internet of Vehicles information and the local perception information acquired by the vehicle-mounted sensors, and use the real-time local perception information to compensate for the Internet of Vehicles information with communication delay; Step 223: Use 5G NR V2X technology to reduce air interface transmission delay; Step 224: Use multiple communication technologies and switch to another link when one link is delayed too much or interrupted, so as to improve the reliability and timeliness of information.

9. The method for constructing a true value integrating high-precision maps and Internet of Vehicles technology according to claim 1, characterized in that: The step 3 includes: Step 31: Use a 3D object detection algorithm or a virtual private network to project the front view into a top-view coordinate system, decode and output the lane instance ID and travel direction vector, and simplify the lane polygon using an iterative adaptive point algorithm. Step 32: Add traffic light status, road construction zone, and traffic control information provided by the Internet of Vehicles. Step 33: embed priority rules and perform priority sorting to complete the scenario construction.

10. A system for implementing the truth value construction method integrating high-precision map and Internet of Vehicles technology according to claims 1-8, characterized in that: The system comprises: The information processing module is used to obtain high-precision map information and vehicle network communication information. It uses real-time positioning and mapping technology to align the dynamic object data of the vehicle network communication information with the high-precision map coordinate system, compensate for the delay of the vehicle network information, ensure the consistency of data timing, and thus complete data preprocessing; The scenario modeling module is connected to the information processing module and is used to generate lane-level drivable areas and traffic rule constraints based on high-precision map information data. It also adds real-time traffic participant trajectories and events provided by the Internet of Vehicles, embeds traffic light status and priority rules, and completes scenario modeling. The truth value generation module is connected to the information processing module and the scenario modeling module respectively. It is used to generate truth value data through the constructed scenario model after preprocessing data to support the multi-dimensional performance evaluation of the autonomous driving algorithm.

Images