Electric bicycle track data road occupation behavior identification method based on high-precision map
Through the high-precision map-based electric bicycle track data identification method, the accuracy and applicability of the identification of illegal occupation of motor vehicle lanes in the prior art are solved, fast and accurate identification of illegal behaviors and all-weather monitoring within the city, and strong traffic safety management technical support is provided.
Patent Information
- Application Number
- CN202510375966.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-27
- Publication Date
- 2025-06-27
AI Technical Summary
The prior art is difficult to accurately identify the illegal occupation of motor vehicle lanes by electric bicycles, resulting in increased monitoring accuracy and implementation difficulty, and lack of technical solutions that can comprehensively and accurately identify the illegal occupation of motor vehicle lanes by electric bicycles.
The electric bicycle track data identification method based on high-precision maps is adopted to identify illegal road occupation behaviors through technical solutions of data cleaning, map matching, behavior identification and risk quantification, and use emerging data platforms such as shared electric bicycles to achieve all-weather monitoring within the city.
It quickly and accurately identify the illegal road occupation behavior of electric bicycles, improves the accuracy and applicability of monitoring results, and can provide strong technical support for traffic safety management independently of traditional monitoring methods.
Smart Images
Figure CN120220408A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of intelligent transportation information processing, and more specifically, it is designed to provide a method for identifying the lane occupation behavior of electric bicycle trajectory data based on a high-precision map. Background Art
[0002] In recent years, due to its characteristics such as fast speed, convenient operation, and low cost, electric bicycles have become the first choice for citizens' short-distance travel. However, with the rapid increase in the number of electric bicycles, new challenges have been brought to traffic management, especially the frequent occurrence of traffic accidents. The occurrence of these accidents is often directly related to the bad driving habits of riders. Therefore, accurately identifying these behaviors is of great significance for improving the traffic management efficiency of electric bicycles and ensuring road safety.
[0003] Among many risky riding behaviors, the illegal behavior of occupying the motor vehicle lane (hereinafter referred to as illegal lane occupation behavior) has been used as a risk quantification index to replace traffic accidents and conflict risks. The key to identifying the illegal behavior of occupying the motor vehicle lane lies in accurately monitoring the lane occupation situation. Currently, the monitoring methods for obtaining the illegal lane occupation behavior of electric bicycles mainly rely on technologies such as road surface monitoring, image recognition, and GPS positioning. However, due to factors such as road conditions, the behavior patterns of different drivers, and data collection frequencies, the recognition results of illegal lane occupation behaviors based on trajectory data vary greatly, resulting in an increase in monitoring accuracy and implementation difficulty. Therefore, an effective technical means is needed that can simultaneously identify illegal behaviors under different road conditions and behavior patterns and improve the accuracy and applicability of monitoring results. Currently, there is no set of technologies that can comprehensively and accurately identify various illegal lane occupation behaviors of electric bicycles and provide an efficient solution. Summary of the Invention
[0004] The technical problem to be solved by the present invention is to propose a method for identifying illegal lane occupation behaviors based on electric bicycle trajectory data in view of the above deficiencies of the prior art. The method is characterized in that it does not rely on road surface monitoring or image recognition, but uses electric bicycle trajectory data, and through an integrated technical solution of "data cleaning - map matching - behavior recognition - risk quantification", it can quickly and accurately identify illegal lane occupation behaviors. In addition, this method can utilize emerging data platforms such as shared electric bicycles and food delivery electric bicycles, independent of traditional monitoring means, to achieve all-weather monitoring of illegal lane occupation behaviors within the city, quantify the road risk level, and thus provide strong technical support for traffic safety management.
[0005] In order to achieve the above objectives, the present invention adopts the following technical solutions:
[0006] Step 1: Obtain and clean electric bicycle trajectory data, mainly cleaning the trajectory data with missing and drifting data;
[0007] Step 2: Map matching to accurately map the trajectory chain to the urban road network;
[0008] Step 3: Use the trajectory data to identify illegal occupation of roadways;
[0009] Step 4: Division of road section risk levels.
[0010] According to an illegal occupation of roadway behavior recognition method based on electric bicycle trajectory data provided by the present invention, the step 1 of obtaining and cleaning electric bicycle trajectory data includes:
[0011] Step 1.1: The trajectory data of the electric bicycle is derived from the high-precision desensitized data provided by the shared electric bicycle operator, mainly including order numbers, trajectory timestamps, and longitude and latitude coordinates.
[0012] Step 1.2: Use the box plot method to clean the missing points and drift points in the trajectory data. Missing points and drift points refer to the abnormal trajectory data caused by factors such as GPS sensor failures or urban high-rise building blockages during the operation of the electric bicycle. Missing points are manifested as data loss, while drift points are manifested as trajectory points deviating from the true position with a large abnormal deviation.
[0013] According to an illegal occupation of roadway behavior recognition method based on electric bicycle trajectory data provided by the present invention, the step 2 of map matching includes:
[0014] Step 2.1: Prepare the high-precision map interface.
[0015] Step 2.2: Use the "whether inside the intersection" interface of the high-precision map to obtain the trajectory points on the road section for subsequent determination.
[0016] Step 2.3: Use the "query lane information according to longitude and latitude" interface of the high-precision map. After each trajectory point passes through this interface, the lane number attribute, i.e., laneId, will be added.
[0017] Step 2.4: Use the "obtain corresponding lane line data according to lane id" interface of the high-precision map. After each trajectory point passes through this interface, the lane length length and the topological point information points of the rightmost lane boundary line rmkg will be added, and some trajectory points will add the right adjacent lane number, i.e., rfward_id.
[0018] According to an illegal occupation of roadway behavior recognition method based on high-precision map electric bicycle trajectory data provided by the present invention, the step 3 of illegal occupation of roadway behavior recognition includes:
[0019] Step 3.1: Obtain the trajectory points on the road section and arrange the trajectory points in the trajectory sub-chain in ascending order of time.
[0020] Step 3.2: Obtain the lane number laneId of the trajectory sub-chain, and for each trajectory sub-chain T with different laneIds rml Add a new iteration pointer T lane , and check the lane number of the right adjacent lane of the trajectory point.
[0021] Step 3.3: Check the trajectory points without a right adjacent lane number, and obtain the topological point information points of the rightmost lane boundary line rmkg, which is used as the line segment information of the motor-vehicle and non-motor-vehicle separation line for this section.
[0022] Step 3.4: Calculate the distance from each trajectory point to the projection point on the rightmost lane boundary line. If the distance is less than the threshold of 0.3m, it is considered that there is no illegal occupation of the motor vehicle lane.
[0023] Step 3.5: Extract the starting and ending coordinates of the motor-vehicle and non-motor-vehicle separation line sub-segment, and judge the running direction LD of the projection points A and B of the trajectory points a and b on the motor-vehicle and non-motor-vehicle separation line AB , that is, whether it is going with the traffic or against the traffic.
[0024] Step 3.6: Determine whether the trajectory point illegally occupies the motor vehicle lane by calculating the cross product of vectors. This process uses the property of the cross product of vectors to judge the relative position of the point and the line segment, so as to determine whether there is an illegal act. Define the vector representing the projection point P A to P B vector, and the vector is the vector from the projection point P A to the trajectory point P a . Calculate the two-dimensional cross product of these two vectors, and multiply the result by the running direction LD of the trajectory point AB . If the result of the cross product is negative, it means that the trajectory point a is located on the non-motor vehicle lane, so RML a returns 0; if the result is positive, it means that the trajectory point a illegally occupies the motor vehicle lane, and RML a returns 1.
[0025] Step 3.7: Evaluate the total length of the illegally occupied lane trajectories on each section, and judge whether it is determined as an illegal lane occupation behavior based on this length. Calculate the total length L lane of the trajectory segments with RML a equal to 1 for each T rml . Compare the length L rml of the illegally occupied lane trajectory segment with the lane length L lane , and calculate the proportion of the illegally occupied lane length. If the proportion of the illegally occupied lane length reaches or exceeds 70%, the illegal lane occupation behavior RML is recorded as 1; if it is less than 70%, then RML is recorded as 0.
[0026] A method for identifying illegal lane occupation behavior based on electric bicycle trajectory data of high-precision map provided by the present invention, wherein the division of road section risk level in step 4 includes:
[0027] Step 4.1: Based on the trajectory data of shared electric bicycles, count the occurrence frequency of illegal lane occupation behavior on the test road section within one day, and divide the severity of illegal lane occupation behavior with a suitable frequency pane, and display the risk level of the test road section based on the high-precision map platform. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1 It is a schematic flow chart of a method for identifying illegal lane occupation behavior based on electric bicycle trajectory data of high-precision map;
[0029] Figure 2 It is a schematic diagram of a test road section;
[0030] Figure 3 It is a schematic diagram of a trajectory drift point;
[0031] Figure 4 It is a schematic diagram of missing trajectory points;
[0032] Figure 5 It is a schematic diagram of illegal lane occupation behavior;
[0033] Figure 6 It is a schematic diagram of the determination result of illegal lane occupation behavior on the test road section;
[0034] Figure 7 It is a schematic diagram of the frequency distribution of illegal lane occupation behavior on the test road section. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0035] To make the objectives, technical solutions and advantages of the present invention clearer and more definite, the following further describes the present invention in detail with reference to the accompanying drawings and by way of examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.
[0036] Please refer to the attached Figure 1 , the present invention provides a method for identifying illegal lane occupation behavior based on electric bicycle trajectory data of high-precision map, including the following steps:
[0037] Step 1: Obtain and clean the electric bicycle trajectory data, mainly cleaning the trajectory data with missing and drifting data;
[0038] Step 2: Map matching, matching the road network attributes to each trajectory point;
[0039] Step 3: Use the trajectory data to determine the illegal lane occupation behavior;
[0040] Step 4: Division of road section risk level.
[0041] The following gives application examples of the present invention.
[0042] For example Figure 2 , the trajectory data of shared electric bicycles on the test section in Yizhuang area, Daxing District, Beijing within one day is used to identify illegal occupation of road behavior.
[0043] (1) Acquisition and cleaning of electric bicycle trajectory data
[0044] 1) The electric bicycle trajectory data is sourced from the de-identified dataset of the shared electric bicycle operator and includes order numbers, trajectory timestamps, and longitude and latitude coordinates.
[0045] 2) Sort the n trajectory points in the trajectory chain in ascending order of time, define the trajectory lock-on and unlock points, and assume the current trajectory point and the previous trajectory point are P i and P i-1 respectively. Calculate the distance d i , speed v i-1 , and time t i-1,i between P i-1,i and P i-1,i .
[0046] 3) Using the box plot method, a data cleaning strategy integrating "time threshold + speed threshold" is constructed to accurately identify and remove abnormal points in the data, including missing points and drift points. In the box plot, when the value is lower than Q L - 1.5IQR or higher than Q U + 1.5IQR, it will be identified as an outlier, where Q L and Q U represent the lower quartile and upper quartile respectively, and IQR is the interquartile range (IQR = Q U - Q L ). As shown in Appendix Figure 3 , Figure 4 , for drift points, when d i-1,i > v thr t i-1,i , delete the trajectory point P i . Among them, v thr is the upper limit value of the allowable speed. For missing points, when t i-1,i > t thr , truncate the original trajectory chain to generate a new trajectory sub-chain. Among them, v thr is the upper limit value of the allowable time. For the trajectory sub-chain, when the number of trajectory points continuously exceeds 10, retain it as the trajectory sub-chain.
[0047] (2) Map matching
[0048] 1) Road network data is obtained through multiple interfaces of the high-precision map, including interfaces such as "whether inside the intersection", "query lane information according to longitude and latitude", and "obtain corresponding lane line data according to lane id".
[0049] 2) Arrange the trajectory sub-chain T in ascending order of time series mm and the trajectory points in the sub-chain.
[0050] 3) The present invention adopts a multi-interface collaborative processing method based on the high-precision map for map matching. First, use the "whether inside the intersection" interface to screen out the trajectory points on the road section, providing basic data for subsequent determination. Then, call the "query lane information according to longitude and latitude" interface to match and add the lane number attribute (laneId) to each trajectory point, realizing the precise association between the trajectory point and the lane. Then, through the "obtain corresponding lane line data according to lane id" interface, further enrich the trajectory point information, making it include the lane length (length), the topological point information (points) of the rightmost lane boundary line (rmkg), and some trajectory points will also add the right adjacent lane number (rfward_id).
[0051] (3) Illegal lane occupation behavior recognition
[0052] 1) Sort the n rml trajectory points in the trajectory sub-chain T in ascending order of time, and define the current trajectory point and the next trajectory point as P rml and P a respectively. b
[0053] 2) Obtain the lane number laneId of the trajectory sub-chain, and add a new iteration pointer T rml to each trajectory sub-chain T with different laneIds lane . Check whether the right adjacent lane number of the trajectory point P lane exists for each T i . If it exists, identify P i as an illegal lane occupation point, and record RML i as 1.
[0054] 3) Check the points where the right adjacent lane number of the trajectory point P i does not exist, and obtain the topological point information L id of the rightmost lane boundary line rmkg as the line segment information of the motor vehicle and non-motor vehicle separation line on this road section.
[0055] 4) First, calculate the projection points P a and P b of each trajectory point P A and P B onto the rightmost lane boundary lineIf the distance is less than the threshold of 0.3 m, it is considered that there is no illegal occupation of the motor vehicle lane. Secondly, judge the running direction LD A of the projection points P B in the lane AB . Extract the starting and ending coordinates of the non-motor vehicle and motor vehicle dividing line sub-segments If the projection points P A and P B of two trajectory points are not within the range of any sub-segment, LD AB is recorded as 0, indicating non-occupation of the lane. If the two projection points P A and P B are located in different sub-segment ranges, compare the n in the coordinates of the sub-segments i to determine the running direction. If n A is less than n B , it is a forward movement, and LD AB is recorded as 1; otherwise, it is a reverse movement, and LD AB is recorded as -1. If the two projection points P A and P B are located on the same sub-segment, the function calculates the distances L A and L B from the two projection points P to the starting point of the sub-segment n1A and L n1B , and compares the relative magnitudes of the distances. If L n1A < L n1B , it is a forward movement, and LD AB is recorded as 1; otherwise, it is a reverse movement, and LD AB is recorded as -1.
[0056] 5) Determine whether the trajectory point illegally occupies the motor vehicle lane by calculating the cross product of vectors. Define the vector representing the vector from the projection point P A to P B , and the vector is the vector from the projection point P A to the trajectory point P a . Calculate the two-dimensional cross product of these two vectors, and multiply the result by the running direction LD AB of the trajectory point. If the result of the cross product is negative, it means that the trajectory point a is on the non-motor vehicle lane, so RML a is recorded as 0; if the result is positive, it means that the trajectory point a illegally occupies the motor vehicle lane, and RML a is recorded as 1.
[0057] 6) Calculate the total length L lane of the trajectory segments where RML a is equal to 1 for each unit of T rml. Compare the length L of the illegal lane occupation trajectory segment rml and the lane length L lane , and calculate the proportion of the illegal lane occupation length. If the proportion of the illegal lane occupation length reaches or exceeds 70%, record it as 1 in the RML field; if it is less than 70%, record it as 0.
[0058] (4) Road section risk level division
[0059] As Figure 5 , Figure 6 shown, based on the trajectory data of shared electric bicycles, count the occurrence frequency of illegal lane occupation behaviors on the test road section within a day, and divide it into three grade levels with appropriate frequency windows, and use three colors of red, orange, and yellow to mark the road section on the GIS map platform to indicate the severity of illegal lane occupation behaviors, so as to observe the risk levels of different road sections, as shown in the appendix Figure 7 shown.
[0060] Matters not covered in this invention are well-known technologies.
[0061] The above embodiments are only used to illustrate the technical concept and features of the present invention, and their purpose is to enable those who are familiar with this technology to understand the content of the present invention and implement it accordingly, and it cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit and essence of the present invention should be covered within the protection scope of the present invention.
Claims
1. A method for identifying road-occupying behavior of electric bicycles based on high-precision maps, characterized in that: By analyzing high-precision map data and trajectory points, the road-occupying behavior of electric bicycles is accurately identified. The method includes the following steps: Step 1: Acquisition and cleaning of electric bicycle trajectory data; Step 2: Map matching, matching high-precision map attributes to each trajectory point or accurately mapping the trajectory chain to the urban road network; Step 3: Use trajectory data to determine the road-occupying behavior; Step 4: Classify road sections into risk levels.
2. According to claim 1, a method for identifying road occupation behavior of electric bicycle trajectory data based on high-precision map is characterized in that: The process of acquiring and cleaning the trajectory data in step 1 is as follows: Step 11: The electric bicycle trajectory data comes from the desensitized data of the shared electric bicycle operator, including the order number, trajectory timestamp, and latitude and longitude coordinates; Step 12: Clean the missing points and drift points in the trajectory data; missing points and drift points refer to the data anomalies of the trajectory points of the vehicle during operation due to factors such as failure of the electric bicycle GPS sensor equipment or obstruction by high-rise buildings in the city, which are manifested as deviations from the actual position or missing data points.
3. The method for identifying road-occupying behavior of electric bicycles based on high-precision maps according to claim 1, characterized in that: The map matching process in step 2 is as follows: Step 21: HD map interface preparation; Step 22: Use the "Is it in the intersection" interface of the high-precision map to obtain the trajectory points on the road section for subsequent judgment; Step 23: Use the "Query lane information based on latitude and longitude" interface of the high-precision map. The lane number attribute will be added to each track point after it passes through this interface; Step 24: Use the "Get corresponding lane line data based on lane ID" interface of the high-precision map. After each trajectory point passes through this interface, the lane length and the topological point information of the rightmost lane edge will be added. Some trajectory points will also have the number of the right adjacent lane added.
4. The method for identifying road-occupying behavior of electric bicycles based on high-precision maps according to claim 1, characterized in that: The process of identifying the sudden speed change behavior in step 3 is as follows: Step 31: Get the trajectory points on the road segment and arrange the trajectory points in the trajectory subchain in ascending time order; Step 32: Get the lane number laneId of the trajectory subchain, and add a new iteration pointer for each trajectory subchain with different laneId, and check the lane number adjacent to the right of the trajectory point; Step 33: Check if there is no trajectory point with the number of the right adjacent lane, obtain the topological point information points of the rightmost lane edge line rmkg, and use it as the line segment information of the vehicle-non-vehicle separation line of the road section; Step 34: Calculate the distance from each trajectory point to the projection point on the rightmost lane edge line. If the distance is less than a threshold, it is considered that there is no illegal occupation of the motor vehicle lane. Step 35: extract the starting and ending coordinates of the non-separating line sub-segment, and determine the running direction of the projection points A and B of the trajectory points a and b on the non-separating line in the lane, that is, whether it is forward or reverse; Step 36: Determine whether the trajectory point illegally occupies the motor vehicle lane by calculating the vector cross product; define vector represents the vector from the projection point to , and vector is the vector from the projection point to the trajectory point; calculate the two-dimensional cross product of these two vectors, and multiply the result by the running direction of the trajectory point. If the result of the cross product is a negative number, it means that the trajectory point a is located on the non-motor vehicle lane, so 0 is returned; if the result is a positive number, it means that the trajectory point a illegally occupies the motor vehicle lane, and 1 is returned; Step 37: Evaluate the total length of the illegal road occupation trajectory on each road section, and determine whether it is an illegal road occupation behavior based on this length; calculate the total length of the trajectory segment equal to 1 for each unit; compare the length of the illegal road occupation trajectory segment and the lane length, and calculate the proportion of the illegal road occupation length; if the proportion of the illegal road occupation length reaches or exceeds 70%, the illegal road occupation behavior is recorded as 1; if it is less than 70%, it is recorded as 0.
5. The method for identifying road-occupying behavior of electric bicycles based on high-precision maps according to claim 1, characterized in that: The process of dividing the road section risk level in step 4 is as follows: Based on the trajectory data of shared electric bicycles, the frequency of illegal road occupation on the test section within a day is counted, and the severity of illegal road occupation is divided into appropriate frequency panes, and the risk level of the test section is displayed based on the high-precision map platform.