Vehicle track correction method and system based on map matching

Abstract

The invention particularly relates to a vehicle track correction method and system based on map matching. According to the vehicle track correction method based on map matching, original vehicle coordinate data are preprocessed and stored in a time sequence database, and duplicate removal processing is carried out; after outliers in the vehicle coordinate data are removed, thinning is carried out; traversing the vehicle coordinate data list, and marking candidate road sections; if the starting point is not the starting point, diffusing all reserved candidate road sections of the previous coordinate point, and calculating the shortest path distance between every two road sections; cost is calculated, when an end point is searched, a candidate road section with the minimum cost is found, and a vehicle track after deviation correction is obtained and displayed on a map. According to the vehicle track correction method based on map matching, track data quality is improved, data credibility and decision support are improved, deep fusion of vehicle track data and map data is realized, cooperative work of subsystems in an intelligent traffic system is promoted, and overall operation efficiency and service level are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of map matching, in particular to a vehicle trajectory correction method and system based on map matching. BACKGROUND

[0002] In the current rapid development of intelligent transportation and Internet of Vehicles, the application of vehicle trajectory information is becoming more and more widespread, such as traffic flow analysis, intelligent driving assistance, etc., and the demand for trajectory accuracy is increasing.

[0003] However, the current trajectory data acquisition mainly relies on GPS positioning, and the positioning accuracy is limited, usually about 3-5 meters, and is affected by many factors, resulting in inaccurate positioning data and abnormal trajectory display on the map. In addition to the accuracy limitation of GPS itself and the error in complex environment, the trajectory of the vehicle may be completely lost or severely deviated in signal shielding areas such as mountainous areas or underground parking lots. These problems not only affect the visualization effect of vehicle trajectory, but also may cause a series of serious consequences, such as misleading driving decisions with incorrect trajectory, increasing the risk of accidents, reducing the efficiency of traffic management, inaccurate traffic flow statistics, and road condition prediction deviation; it will also hinder the development of intelligent transportation system, such as autonomous vehicles relying on accurate trajectory information for path planning and decision making, and trajectory deviation may cause it to be unable to accurately judge the road conditions, affecting the safety and reliability of driving.

[0004] Based on the above problems, the present application provides a vehicle trajectory correction method and system based on map matching, aiming to correct the vehicle driving trajectory through technical innovation, improve the quality of trajectory data, and make the data more complete, continuous and real. SUMMARY

[0005] The present application provides a simple and efficient vehicle trajectory correction method based on map matching to overcome the defects of the prior art.

[0006] The present application is realized by the following technical solutions:

[0007] A vehicle trajectory correction method and system based on map matching, characterized by comprising the following steps:

[0008] Step S1, preprocessing the original vehicle coordinate data and storing it in a time series database;

[0009] Step S2, obtaining the vehicle coordinate data list from the time series database and performing de-duplication processing;

[0010] Step S3, removing outliers in the vehicle coordinate data; outliers refer to coordinate points that differ greatly from other surrounding coordinate points and do not conform to the normal trajectory pattern;

[0011] Step S4, thinning the vehicle coordinate data;

[0012] Step S5, traversing the vehicle coordinate data list, for the currently traversed coordinate point, obtaining the road segment points within a circle of 200 meters, and marking the candidate road segments;

[0013] Step S6, if not the starting point, diffusing all the reserved candidate road segments of the last coordinate point, and calculating the shortest path distance between each two road segments;

[0014] Step S7, calculating the cost, which is the sum of the candidate road segment cost of the last coordinate point, the transition cost to the candidate road segment corresponding to the current coordinate point, and the observation cost of the current coordinate point;

[0015] Step S8, when searching to the end point, finding the candidate road segment with the minimum cost, and then tracing back in turn to find all the coordinate points passed when calculating the minimum cost, obtaining the actual road coordinate point list data of the matched map, i.e., the vehicle trajectory after rectification;

[0016] Step S9, displaying the vehicle trajectory after rectification on the map.

[0017] In the step S1, the original vehicle coordinate position data includes license plate number, longitude, latitude, positioning time, direction angle, and speed.

[0018] In the step S4, the criteria for thinning out the vehicle coordinate data are as follows:

[0019] 1) Reserving one vehicle coordinate point within a distance of 50 meters;

[0020] 2) Reserving the vehicle coordinate starting point, and not using the direction information of the starting point;

[0021] 3) Reserving the vehicle coordinate end point, and not using the direction information of the end point;

[0022] 4) Reserving the vehicle coordinate point with a direction angle exceeding a self-defined threshold.

[0023] In the step S7, the transition distance difference is used as the transition cost, i.e., the absolute value of the difference between the distance difference of two adjacent actual vehicle coordinate points and the distance difference of the candidate road segment coordinate points of the two coordinate points;

[0024] The observation cost is the weighted average of the following three cost factors: the distance from the actual coordinate point to the candidate road segment, the difference between the direction angle of the actual coordinate point and the direction angle of the candidate road segment, and the difference between the speed of the actual coordinate point and the speed limit value of the candidate road segment.

[0025] A vehicle trajectory rectification system based on map matching is used to implement the above method, comprising:

[0026] A preprocessing module is responsible for preprocessing original vehicle coordinate data and storing into a time-series database;

[0027] A de-duplication module is responsible for obtaining a vehicle coordinate data list from the time-series database and performing de-duplication processing;

[0028] A wild point removal module is responsible for removing wild points in the vehicle coordinate data; the wild point refers to a coordinate point that is too different from other surrounding coordinate points and does not conform to a normal trajectory mode;

[0029] A thinning module is responsible for thinning the vehicle coordinate data;

[0030] A candidate road segment marking module is responsible for traversing the vehicle coordinate data list, obtaining road segment points within a radius of 200 meters for a currently traversed coordinate point, and marking a candidate road segment;

[0031] A path distance calculation module is responsible for diffusing all retained candidate road segments of a previous coordinate point for a coordinate point that is not a starting point, and calculating a shortest path distance between each two road segments;

[0032] A cost calculation module is responsible for calculating a cost, i.e., a sum of a candidate road segment cost of a previous coordinate point, a transition cost to a candidate road segment corresponding to a current coordinate point, and an observation cost of the current coordinate point;

[0033] A trajectory deviation correction module is responsible for finding a candidate road segment with a minimum cost when a terminal point is searched, and then tracing back in sequence to find all coordinate points passed when the minimum cost is calculated, to obtain actual road coordinate point list data of a matched map, i.e., a vehicle trajectory after deviation correction;

[0034] A deviation-corrected trajectory display module is responsible for displaying the deviation-corrected vehicle trajectory on a map.

[0035] The original vehicle coordinate position data includes a license plate number, a longitude, a latitude, a positioning time, a direction angle, and a speed.

[0036] The thinning module has a criterion for thinning the vehicle coordinate data, as follows:

[0037] 1) One vehicle coordinate point is retained at a distance of 50 meters;

[0038] 2) A vehicle coordinate starting point is retained, and direction information of the starting point is not used;

[0039] 3) A vehicle coordinate terminal point is retained, and direction information of the terminal point is not used;

[0040] 4) A vehicle coordinate point with a direction angle exceeding a self-defined threshold is retained.

[0041] The conversion distance difference is used as the conversion cost, that is, the absolute value of the difference between the actual distance difference between two adjacent vehicle coordinate points and the distance difference between the two coordinate points and the candidate road segment coordinate points;

[0042] The observation cost is a weighted average of the following three cost factors: the distance between the actual coordinate point and the candidate road segment, the difference between the actual coordinate point direction angle and the candidate road segment direction angle, and the difference between the actual coordinate point speed and the candidate road segment speed limit value.

[0043] A vehicle trajectory correction device based on map matching, characterized by comprising a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to realize the method steps described above.

[0044] A readable storage medium, characterized by: the readable storage medium has a computer program stored thereon, and the computer program is executed by a processor to realize the method steps described above.

[0045] The beneficial effects of the present application are: the vehicle trajectory correction method based on map matching not only improves the quality of trajectory data, makes the data more complete, continuous and real, improves the data credibility and decision support, provides reliable basis for road planning, traffic flow analysis, accident risk assessment and other traffic management departments, and helps scientific decision-making; at the same time, it can also realize the deep fusion of vehicle trajectory data and map data, promote the collaborative work of each subsystem in the intelligent transportation system, and improve the overall operation efficiency and service level. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings described below are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0047] Appendix Figure 1 The figure shows the vehicle trajectory correction method based on map matching. DETAILED DESCRIPTION

[0048] In order to make the person skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0049] The vehicle trajectory correction method based on map matching comprises the following steps:

[0050] Step S1, preprocessing the original vehicle coordinate data and storing it in a time series database;

[0051] Step S2, obtaining the vehicle coordinate data list from the time series database and performing de-duplication processing;

[0052] Step S3, removing outliers in the vehicle coordinate data; outliers refer to coordinate points that differ greatly from other surrounding coordinate points and do not conform to the normal trajectory pattern;

[0053] Step S4, thinning the vehicle coordinate data;

[0054] Step S5, traversing the vehicle coordinate data list, obtaining the road segment points within a radius of 200 meters for the current traversed coordinate point, and marking the candidate road segments;

[0055] Step S6, if it is not the starting point, diffusing all the retained candidate road segments of the previous coordinate point, and calculating the shortest path distance between each two road segments;

[0056] Step S7, calculating the cost, which is the sum of the candidate road segment cost of the previous coordinate point, the transition cost to the candidate road segment corresponding to the current coordinate point, and the observation cost of the current coordinate point;

[0057] Step S8, when the end point is searched, the candidate road segment with the minimum cost is found, and then the all coordinate points passed when calculating the minimum cost are found by tracing back in sequence, obtaining the actual road coordinate point list data of the matched map, i.e., the corrected vehicle trajectory;

[0058] Step S9, displaying the corrected vehicle trajectory on the map.

[0059] In step S1, the original vehicle coordinate position data includes license plate number, longitude, latitude, positioning time, direction angle, and speed.

[0060] In step S4, the criteria for thinning the vehicle coordinate data are as follows:

[0061] 1) Keep one vehicle coordinate point every 50 meters;

[0062] 2) Keep the starting point of the vehicle coordinate and do not use the direction information of the starting point;

[0063] 3) Keep the end point of the vehicle coordinate and do not use the direction information of the end point;

[0064] 4) Keep the vehicle coordinate point whose direction angle exceeds the self-defined threshold.

[0065] The conversion distance difference is used as the conversion cost in the step S7, that is, the absolute value of the difference between the actual distance difference between two adjacent vehicle coordinate points and the distance difference between the two coordinate points and the candidate road segment coordinate points;

[0066] The observation cost is a weighted average of the following three cost factors: the distance between the actual coordinate point and the candidate road segment, the difference between the direction angle of the actual coordinate point and the direction angle of the candidate road segment, and the difference between the speed of the actual coordinate point and the speed limit value of the candidate road segment.

[0067] The vehicle trajectory correction system based on map matching is used to implement the above method and comprises:

[0068] A preprocessing module is responsible for preprocessing the original vehicle coordinate data and storing the same in a time sequence database.

[0069] A de-duplication module is responsible for obtaining a vehicle coordinate data list from the time sequence database and performing de-duplication processing.

[0070] A wild point removal module is responsible for removing wild points in the vehicle coordinate data. A wild point refers to a coordinate point that is too different from other surrounding coordinate points and does not conform to a normal trajectory mode.

[0071] A thinning module is responsible for thinning the vehicle coordinate data.

[0072] A candidate road segment marking module is responsible for traversing the vehicle coordinate data list, obtaining road segment points within a radius of 200 meters for a currently traversed coordinate point, and marking a candidate road segment.

[0073] A path distance calculation module is responsible for diffusing all retained candidate road segments of a previous coordinate point for a coordinate point that is not a starting point and calculating the shortest path distance between each two road segments.

[0074] A cost calculation module is responsible for calculating a cost, that is, the sum of the candidate road segment cost of a previous coordinate point, the conversion cost to a candidate road segment corresponding to a current coordinate point, and the observation cost of the current coordinate point.

[0075] A trajectory correction module is responsible for finding a candidate road segment with the minimum cost when a terminal point is searched, then tracing back in sequence to find all coordinate points that pass through when the minimum cost is calculated, obtaining a list of actual road coordinate points of a matched map, that is, a corrected vehicle trajectory.

[0076] A corrected trajectory display module is responsible for displaying the corrected vehicle trajectory on a map.

[0077] The original vehicle coordinate position data comprises a license plate number, a longitude, a latitude, a positioning time, a direction angle, and a speed.

[0078] The criteria for the vehicle coordinate data sparsification module are as follows:

[0079] 1) Keep 1 vehicle coordinate point every 50 meters;

[0080] 2) Keep the starting point of the vehicle coordinates and do not use the direction information of the starting point;

[0081] 3) Keep the ending point of the vehicle coordinates and do not use the direction information of the ending point;

[0082] 4) Keep the vehicle coordinate point whose direction angle exceeds the self-defined threshold.

[0083] The conversion distance difference is used as the conversion cost, i.e., the absolute value of the difference between the actual distance difference between two adjacent vehicle coordinate points and the distance difference between the candidate road segment coordinate points of the two coordinate points;

[0084] The observation cost is the weighted average of the following three cost factors: the distance from the actual coordinate point to the candidate road segment, the difference between the direction angle of the actual coordinate point and the direction angle of the candidate road segment, and the difference between the speed of the actual coordinate point and the speed limit value of the candidate road segment.

[0085] The map-matching-based vehicle trajectory correction device includes a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the method steps described above when executing the computer program.

[0086] The readable storage medium stores a computer program, and the computer program is executed by the processor to implement the method steps described above.

[0087] Compared with the prior art, the map-matching-based vehicle trajectory correction method has the following characteristics:

[0088] (1) Improving data reliability and decision support: By correcting the trajectory deviation, high-precision and accurate vehicle driving data are provided, which can provide reliable basis for road planning, traffic flow analysis, accident risk assessment, etc. for traffic management departments, and help scientific decision-making.

[0089] (2) Improving the quality of trajectory data: optimizing the original trajectory data, removing noise and errors, making the data more complete, continuous and true, and facilitating subsequent in-depth analysis and application, such as traffic pattern research and driving behavior analysis.

[0090] (3) Helping intelligent transportation system integration: realizing the deep integration of vehicle trajectory data and map data, promoting the collaborative work of each subsystem in the intelligent transportation system, and improving the overall operation efficiency and service level.

[0091] The above-mentioned embodiments are only one of the specific embodiments of the present application, and the common changes and replacements made by the skilled in the art within the technical scheme of the present application should be included in the protection scope of the present application.

Claims

1. A vehicle trajectory correction method and system based on map matching, characterized in that: Includes the following steps: Step S1: Preprocess the raw vehicle coordinate data and store it in the time series database; Step S2: Obtain a list of vehicle coordinate data from the time series database and perform deduplication. Step S3: Remove outliers from the vehicle coordinate data; Step S4: Thin out the vehicle coordinate data; Step S5: Traverse the vehicle coordinate data list, obtain the road segment locations within a 200-meter radius of the currently traversed coordinate point, and mark the candidate road segments. Step S6: If it is not the starting point, then expand all the remaining candidate road segments of the previous coordinate point and calculate the shortest path distance between each pair of road segments. Step S7: Calculate the cost, which is the sum of the candidate road segment cost of the previous coordinate point, the conversion cost of the corresponding candidate road segment to the current coordinate point, and the observation cost of the current coordinate point. Step S8: When the destination is found, find the candidate road segment with the lowest cost, and then trace back in sequence to find all the coordinate points passed when the minimum cost was calculated, and obtain the list of actual road coordinate points matching the map, that is, the vehicle trajectory after correction. Step S9: Display the corrected vehicle trajectory on the map.

2. The vehicle trajectory correction method based on map matching according to claim 1, characterized in that: In step S1, the original vehicle coordinate location data includes license plate number, longitude, latitude, positioning time, direction angle, and speed.

3. The vehicle trajectory correction method based on map matching according to claim 1, characterized in that: In step S4, the criteria for thinning the vehicle coordinate data are as follows: 1) Maintain one vehicle coordinate point every 50 meters; 2) Retain the vehicle's starting coordinates, but do not use the direction information of the starting point; 3) Retain the vehicle's endpoint coordinates, but do not use the endpoint's direction information; 4) Retain vehicle coordinates that exceed a custom threshold.

4. The vehicle trajectory correction method based on map matching according to claim 1, characterized in that: In step S7, the conversion distance difference is used as the conversion cost, which is the absolute value of the difference between the actual distance difference between two adjacent vehicle coordinate points and the distance difference between the candidate road segment coordinate points. The observation cost is a weighted average of the following three cost factors: the distance from the actual coordinate point to the candidate road segment, the difference between the direction angle of the actual coordinate point and the direction angle of the candidate road segment, and the difference between the speed of the actual coordinate point and the speed limit of the candidate road segment.

5. A vehicle trajectory correction system based on map matching, characterized in that: To implement the method according to any one of claims 1 to 4, comprising: The preprocessing module is responsible for preprocessing the raw vehicle coordinate data and storing it in the time-series database; The deduplication module is responsible for obtaining a list of vehicle coordinate data from the time-series database and performing deduplication. The outlier removal module is responsible for removing outliers from vehicle coordinate data. The thinning module is responsible for thinning the vehicle coordinate data. The candidate road segment marking module is responsible for traversing the vehicle coordinate data list, obtaining the road segment locations within a 200-meter radius of the currently traversed coordinate point, and marking the candidate road segment. The path distance calculation module is responsible for expanding all the reserved candidate road segments of the previous coordinate point for coordinate points that are not the starting point, and calculating the shortest path distance between each pair of road segments. The cost calculation module is responsible for calculating the cost, which is the sum of the cost of the candidate road segment at the previous coordinate point, the conversion cost of the corresponding candidate road segment to the current coordinate point, and the observation cost of the current coordinate point. The trajectory correction module is responsible for finding the candidate road segment with the lowest cost when the search reaches the destination, and then tracing back sequentially to find all the coordinate points passed when the minimum cost was calculated, and obtaining the list of actual road coordinate points matching the map, which is the corrected vehicle trajectory. The trajectory correction display module is responsible for displaying the corrected vehicle trajectory on the map.

6. The vehicle trajectory correction system based on map matching according to claim 5, characterized in that: The original vehicle coordinate location data includes license plate number, longitude, latitude, positioning time, direction angle, and speed.

7. The vehicle trajectory correction method based on map matching according to claim 5, characterized in that: The criteria for the thinning module to thin the vehicle coordinate data are as follows: 1) Maintain one vehicle coordinate point every 50 meters; 2) Retain the vehicle's starting coordinates, but do not use the direction information of the starting point; 3) Retain the vehicle's endpoint coordinates, but do not use the endpoint's direction information; 4) Retain vehicle coordinates that exceed a custom threshold.

8. The vehicle trajectory correction system based on map matching according to claim 5, characterized in that: The conversion distance difference is used as the conversion cost, which is the absolute value of the difference between the actual distance difference between two adjacent vehicle coordinate points and the distance difference between the candidate road segment coordinate points. The observation cost is a weighted average of the following three cost factors: the distance from the actual coordinate point to the candidate road segment, the difference between the direction angle of the actual coordinate point and the direction angle of the candidate road segment, and the difference between the speed of the actual coordinate point and the speed limit of the candidate road segment.

9. A vehicle trajectory correction device based on map matching, characterized in that: It includes a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to implement the steps of the method as described in any one of claims 1 to 4.

10. A readable storage medium, characterized in that: The readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method as described in any one of claims 1 to 4.

Images