Map file generation method and system based on RTK historical track information

Abstract

The invention relates to a map file generation method and system based on RTK historical track information, and the method comprises the steps: collecting the RTK historical track information of a vehicle, and carrying out the preprocessing, specifically, carrying out the denoising processing of the RTK historical track information, carrying out the segmentation processing of the denoised track data according to the continuity of the track and the road characteristics, and carrying out the segmentation processing of the segmented data; performing alignment processing on the plurality of sections of tracks after segmentation processing to obtain a plurality of pieces of track data; clustering the obtained multiple pieces of trajectory data by using a trajectory clustering algorithm, and extracting an initial topological structure of a road center line; clustering results are fused through a track fusion technology, and a road center line topology is obtained; analyzing road network structure information of the road based on the road center line topology; and converting into a map file according to the road center line topology and the corresponding road network structure information. Compared with the prior art, the method has the advantages of high flexibility, high efficiency, high precision and the like, and can meet the requirements of an automatic driving system on a high-precision map in a limited scene.

Description

Technical Field

[0001] The present invention relates to the technical field of map generation, and in particular, to a method and system for generating a map file based on RTK historical trajectory information. Background Art

[0002] High-precision maps are an important part of autonomous driving systems, and their accuracy and integrity directly affect the positioning, path planning, and decision-making control of autonomous vehicles. Existing methods for generating high-precision maps usually rely on environmental data collected by sensors such as lidar and cameras, combined with manual annotation or automated processing. However, these methods have the following problems:

[0003] 1. High cost: The hardware costs of lidar and high-performance cameras are relatively high, and data processing is complex.

[0004] 2. Low efficiency: The map generation process requires a large amount of manual participation, especially in complex road network environments with multiple lanes and multiple directions.

[0005] 3. Poor real-time performance: Existing methods are difficult to quickly generate high-precision maps covering large areas. Summary of the Invention

[0006] The purpose of the present invention is to provide a method and system for generating a map file based on RTK historical trajectory information to overcome the defects of the existing high-precision map generation methods, such as high cost and low efficiency.

[0007] The purpose of the present invention can be achieved by the following technical solutions:

[0008] A method for generating a map file based on RTK historical trajectory information includes the following steps:

[0009] Collect the RTK historical trajectory information of a vehicle;

[0010] Preprocess the collected RTK historical trajectory information. The preprocessing process includes: denoising the RTK historical trajectory information, segmenting the trajectory data obtained after denoising according to the continuity of the trajectory and road features, aligning the multiple segments of the segmented trajectory to obtain multiple pieces of trajectory data;

[0011] Use a trajectory clustering algorithm to cluster the multiple pieces of trajectory data obtained, and extract the initial topological structure of the road centerline; and fuse the clustering results through a trajectory fusion technology to obtain the road centerline topology;

[0012] Analyze the road network structure information of the road based on the road centerline topology;

[0013] Convert the road centerline topology and the corresponding road network structure information into a map file.

[0014] Further, the denoising process is specifically as follows:

[0015] Filter the abnormal points in the RTK historical trajectory information to remove the noise data.

[0016] Further, the trajectory fusion technology is the weighted average method or the curve fitting method.

[0017] Further, the trajectory clustering algorithm is the DBSCAN algorithm.

[0018] Further, the road network structure information of the road includes the number of lanes, the lane width, and the intersection topological relationship.

[0019] Further, the method uses the ScenarioGeneration tool to convert the road centerline topology and the corresponding road network structure information into a map file conforming to the OpenDRIVE format.

[0020] Further, the method further includes:

[0021] Load the generated map file using a simulation tool;

[0022] In the simulation environment, verify whether the vehicle outline does not exceed the lane lines of the generated map when the vehicle travels along the trajectory line. If it exceeds, further optimize the map file.

[0023] The present invention also provides a map file generation system based on RTK historical trajectory information, including:

[0024] A data acquisition module for acquiring the RTK historical trajectory information of the vehicle;

[0025] A preprocessing module for preprocessing the acquired RTK historical trajectory information. This preprocessing process includes: performing denoising processing on the RTK historical trajectory information, segmenting the trajectory data obtained after denoising according to the continuity of the trajectory and road characteristics, and performing alignment processing on the multiple segments of trajectories after segmentation processing to obtain multiple trajectory data;

[0026] A trajectory clustering and fusion module for clustering the obtained multiple trajectory data using a trajectory clustering algorithm to extract the initial topological structure of the road centerline; and fusing the clustering results through a trajectory fusion technology to obtain the road centerline topology;

[0027] A road network structure analysis module for analyzing the road network structure information of the road based on the road centerline topology;

[0028] A map file conversion module for converting into a map file according to the road centerline topology and the corresponding road network structure information.

[0029] Furthermore, the denoising process of the preprocessing module is specifically as follows:

[0030] Filter the abnormal points in the RTK historical trajectory information to remove the noise data;

[0031] The trajectory fusion technology is the weighted average and curve fitting method;

[0032] The trajectory clustering algorithm is the DBSCAN algorithm;

[0033] The road network structure information of the road includes the number of lanes, lane width, and intersection topological relationship;

[0034] The map file conversion module uses the ScenarioGeneration tool to convert the road centerline topology and the corresponding road network structure information into a map file that conforms to the OpenDRIVE format.

[0035] Furthermore, the system further includes:

[0036] A simulation verification module, which is used to load the generated map file by using a simulation tool; in the simulation environment, verify whether the vehicle outline will not exceed the lane lines of the generated map when the vehicle travels along the trajectory line. If it exceeds, further optimize the map file.

[0037] Compared with the prior art, the present invention has the following advantages:

[0038] (1) Low cost: The present invention collects the RTK historical trajectory information of the vehicle, uses the trajectory clustering and fusion technology to generate the road centerline topology, and generates a high-precision map file that conforms to the OpenDRIVE format based on the road centerline topology. Only an RTK device is required to complete the trajectory data collection, without relying on expensive lidar or cameras.

[0039] (2) High efficiency: Through the trajectory clustering and fusion technology, the road centerline topology is quickly generated, significantly improving the map generation efficiency.

[0040] (3) High precision: Utilize the centimeter-level positioning accuracy of the RTK technology to ensure that the generated high-precision map file meets the requirements of autonomous driving.

[0041] (4) Verifiability: Verify the accuracy and applicability of the map file through a simulation tool to ensure that the generated map file can meet the actual application requirements.

[0042] (5) Suitable for the deployment of autonomous driving solutions in restricted areas: In mines and short-distance transfer scenarios, the objective road network conditions are relatively simple, and the traffic signs and markings involved are limited. The map file generation method provided by the present invention can be used to quickly carry out business deployment. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic flow chart of a map file generation method based on RTK historical trajectory information provided in an embodiment of the present invention;

[0044] Figure 2 It is a schematic diagram of the acquisition result of RTK trajectory data provided in an embodiment of the present invention;

[0045] Figure 3 It is a schematic diagram of the trajectory clustering and fusion result provided in an embodiment of the present invention;

[0046] Figure 4 It is a schematic diagram of the road network structure analysis result provided in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations.

[0048] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0049] It should be noted that: similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0050] Embodiment 1

[0051] As Figure 1 shown, this embodiment provides a map file generation method based on RTK historical trajectory information, including the following steps:

[0052] S1: Collect the RTK historical trajectory information of the vehicle;

[0053] S2: Preprocess the collected RTK historical trajectory information. This preprocessing process includes: denoising the RTK historical trajectory information, segmenting the trajectory data obtained after denoising according to the continuity of the trajectory and road features, and aligning the multiple segments of the segmented trajectory to obtain multiple trajectory data;

[0054] S3: Use the trajectory clustering algorithm to cluster the obtained multiple trajectory data, and extract the initial topological structure of the road centerline; and fuse the clustering results through the trajectory fusion technology to obtain the road centerline topology;

[0055] S4: Analyze the road network structure information based on the road centerline topology;

[0056] S5: Convert the road centerline topology and the corresponding road network structure information into a map file.

[0057] Specifically, step S1 realizes the collection of RTK trajectory data:

[0058] On the premise of stable RTK positioning of the vehicle, collect the RTK historical trajectory information of the vehicle, covering the road trajectory data of multiple lanes and multiple directions, as Figure 2 shown.

[0059] That is:

[0060] 1. On the premise of stable RTK positioning of the vehicle, use the RTK device to collect the historical trajectory information of the vehicle.

[0061] 2. During the collection process, cover the road trajectory information of multiple lanes and multiple directions to ensure the comprehensiveness of the trajectory data.

[0062] 3. The collected data includes information such as timestamp, longitude and latitude, heading angle, and speed.

[0063] Step S2 realizes the preprocessing of trajectory data, specifically including:

[0064] 1. Denoising processing: Filter the abnormal points in the trajectory data to remove the noise data.

[0065] 2. Trajectory segmentation: Segment the trajectory data according to the continuity of the trajectory and road features.

[0066] 3. Trajectory alignment: Align multiple trajectories to ensure the spatial consistency of the trajectory data.

[0067] Step S3 realizes trajectory clustering and fusion, specifically including:

[0068] Trajectory clustering: Use the trajectory clustering algorithm to cluster multiple trajectories and extract the initial topological structure of the road centerline;

[0069] The DBSCAN (Density-Based Spatial Clustering of Applications with Noise) or other clustering algorithms can be used to cluster multiple trajectories and extract the initial topology of the road centerline.

[0070] Trajectory fusion: Optimize the centerline through trajectory fusion technology to generate a high-precision road centerline topology.

[0071] The clustering results can be fused by methods such as weighted average and curve fitting to generate a high-precision road centerline topology, as Figure 3 shown.

[0072] The specific process of road network structure analysis in step S4 is as follows:

[0073] 1. Based on the road centerline topology, analyze the road network structure, including the number of lanes, lane width, intersection topological relationship, etc., as Figure 4 shown.

[0074] 2. Extract the key features of the road network structure to provide input data for the generation of high-precision map files.

[0075] The specific process of realizing the generation of high-precision map files in step S5 is as follows:

[0076] 1. Use tools similar to ScenarioGeneration to convert the road centerline topology and road network structure information into high-precision map files (xodr files) that conform to the OpenDRIVE format.

[0077] 2. During the generation process, ensure that the format and content of the map file conform to the OpenDRIVE standard.

[0078] Preferably, the method further includes:

[0079] Use a simulation tool to load the generated map file;

[0080] In the simulation environment, verify that when the vehicle travels along the trajectory line, the outer contour of the vehicle does not exceed the lane lines of the generated map. If it exceeds, further optimize the map file.

[0081] Equivalent to:

[0082] 1. Use simulation tools (such as CARLA, SUMO, etc.) to load the generated high-precision map file.

[0083] 2. In the simulation environment, verify that when the vehicle travels along the trajectory line, its outer contour does not exceed the lane lines of the generated map.

[0084] 3. Further optimize the map file based on the verification results to ensure its accuracy and applicability.

[0085] The above is the introduction of the method embodiments. The following further illustrates the solution of the present invention through system embodiments.

[0086] This embodiment also provides a map file generation system based on RTK historical trajectory information, including:

[0087] A data acquisition module for acquiring RTK historical trajectory information of vehicles;

[0088] A preprocessing module for preprocessing the acquired RTK historical trajectory information. This preprocessing process includes: denoising the RTK historical trajectory information, segmenting the trajectory data obtained after denoising according to the continuity of the trajectory and road features, aligning the multiple segments of trajectories after segmentation processing to obtain multiple trajectory data;

[0089] A trajectory clustering and fusion module for clustering the obtained multiple trajectory data using a trajectory clustering algorithm to extract the initial topological structure of the road centerline; and fusing the clustering results through a trajectory fusion technology to obtain the road centerline topology;

[0090] A road network structure analysis module for analyzing the road network structure information of the road based on the road centerline topology;

[0091] A map file conversion module for converting into a map file according to the road centerline topology and the corresponding road network structure information.

[0092] The denoising process of the preprocessing module is specifically:

[0093] Filtering the abnormal points in the RTK historical trajectory information to remove the noise data;

[0094] The trajectory fusion technology is the weighted average method or the curve fitting method;

[0095] The trajectory clustering algorithm is the DBSCAN algorithm;

[0096] The road network structure information of the road includes the number of lanes, lane width, and intersection topological relationship;

[0097] The map file conversion module uses the ScenarioGeneration tool to convert the road centerline topology and the corresponding road network structure information into a map file conforming to the OpenDRIVE format.

[0098] Preferably, the system further includes:

[0099] A simulation verification module is used to load the generated map file by means of a simulation tool; in the simulation environment, it verifies whether the outer contour of the vehicle will not exceed the lane lines of the generated map when the vehicle travels along the trajectory line. If it exceeds, the map file is further optimized.

[0100] It should be noted that the specific content and beneficial effects of the system of the present application can be referred to the above method embodiments and will not be elaborated here.

[0101] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations according to the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field of the present invention based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art should fall within the protection scope determined by the claims.

Claims

1. A method for generating a map file based on RTK historical trajectory information, characterized in that: The following steps are involved: Collect the vehicle's RTK historical trajectory information; Preprocessing the collected RTK historical trajectory information includes: denoising the RTK historical trajectory information, segmenting the trajectory data obtained after denoising according to the continuity of the trajectory and the road characteristics, aligning the multiple segments of the segmented trajectory to obtain multiple trajectory data; The trajectory clustering algorithm is used to cluster the multiple trajectory data obtained to extract the initial topological structure of the road centerline; and the clustering results are fused through the trajectory fusion technology to obtain the road centerline topology; Analyzing the road network structure information of the road based on the road centerline topology; The road centerline topology and the corresponding road network structure information are converted into a map file.

2. A method for generating a map file based on RTK historical trajectory information according to claim 1, characterized in that: The denoising process is specifically as follows: Filter outliers in RTK historical trajectory information to remove noise data.

3. The method for generating a map file based on RTK historical trajectory information according to claim 1, characterized in that: The trajectory fusion technology is weighted average and curve fitting method.

4. The method for generating a map file based on RTK historical trajectory information according to claim 1, characterized in that: The trajectory clustering algorithm is the DBSCAN algorithm.

5. The method for generating a map file based on RTK historical trajectory information according to claim 1, characterized in that: The road network structure information of the road includes the number of lanes, lane width and intersection topology.

6. The method for generating a map file based on RTK historical trajectory information according to claim 1, characterized in that: The method uses a ScenarioGeneration tool to convert the road centerline topology and corresponding road network structure information into a map file that complies with the OpenDRIVE format.

7. The method for generating a map file based on RTK historical trajectory information according to claim 1, characterized in that: The method further comprises: Use simulation tools to load the generated map file; In the simulation environment, verify whether the outer contour of the vehicle does not exceed the lane line of the generated map when the vehicle travels along the trajectory line. If it exceeds, further optimize the map file.

8. A map file generation system based on RTK historical trajectory information, characterized in that: include: Data acquisition module, used to collect the vehicle's RTK historical trajectory information; A preprocessing module is used to preprocess the collected RTK historical trajectory information. The preprocessing process includes: denoising the RTK historical trajectory information, segmenting the trajectory data obtained after denoising according to the continuity of the trajectory and the road characteristics, and aligning the multiple segments of the segmented trajectory to obtain multiple trajectory data; The trajectory clustering and fusion module is used to cluster the multiple trajectory data obtained by using the trajectory clustering algorithm to extract the initial topological structure of the road centerline; and to fuse the clustering results through the trajectory fusion technology to obtain the road centerline topology; A road network structure analysis module, used to analyze the road network structure information of the road based on the road centerline topology; A map file conversion module is used to convert the road centerline topology and the corresponding road network structure information into a map file.

9. A map file generation system based on RTK historical trajectory information according to claim 8, characterized in that: The denoising process of the preprocessing module is specifically as follows: Filter outliers in RTK historical trajectory information to remove noise data; The trajectory fusion technology is weighted average and curve fitting method; The trajectory clustering algorithm is the DBSCAN algorithm; The road network structure information of the road includes the number of lanes, lane width and intersection topology; The map file conversion module uses the ScenarioGeneration tool to convert the road centerline topology and the corresponding road network structure information into a map file that complies with the OpenDRIVE format.

10. The map file generation system based on RTK historical trajectory information according to claim 8, characterized in that: The system further comprises: The simulation verification module is used to load the generated map file using a simulation tool; in the simulation environment, it is verified whether the outer contour of the vehicle will not exceed the lane line of the generated map when the vehicle travels along the trajectory line. If it exceeds, the map file is further optimized.

Images