A method for compressing vehicle trajectories

Abstract

The present invention discloses a method for compressing vehicle trajectories. The compression ratio of the original trajectory to the compressed trajectory is set as n. Taking the starting point of the original trajectory as the starting point of the compressed trajectory, the original trajectory is divided into a number of original trajectory segments with a length of n, and the number of original trajectory segments is compressed into a number of compressed trajectory segments with a length of 1, so that the similarity between the compressed trajectory composed of the number of compressed trajectory segments and the original trajectory is maximized. The present invention selects the coordinate points of the compressed trajectory with a dynamic measurement standard, which maximally ensures the similarity between the compressed trajectory and the original trajectory, enables the compressed trajectory to always reflect the uniqueness of the original trajectory, and the reduction degree of the compressed trajectory to the original trajectory is very high. On the basis of ensuring the effective compression and storage of the original trajectory, the travel pattern of the vehicle can be accurately judged according to the compressed trajectory, laying a foundation for the actual application of big data in the later stage, and having good market application value.

Description

Technical Field

[0001] The present invention relates to the technical field of big data analysis for intelligent networking in transportation, and specifically refers to a method for compressing vehicle trajectories. Background Art

[0002] With the rapid development of the Internet of Vehicles, in Internet of Vehicles communication, vehicles (such as smart / intelligent cars, intelligent connected vehicles (ICVs)) can continuously send their own driving data to the cloud server during driving. The cloud server stores the driving data of the vehicles and analyzes the travel patterns of vehicle users based on the stored driving data, and plans routes, predicts destinations, and pushes personalized services for vehicle users according to the travel patterns of vehicle users. In the prior art, the following two methods can be used to analyze the travel patterns of vehicle users based on the driving data of vehicles: Method 1: Cluster the starting and ending points of the driving trajectories of vehicles, and analyze the travel patterns of vehicle users according to the start and stop times and start and stop positions of the vehicles corresponding to the driving trajectories. Among them, the driving trajectory includes the travel time and travel route of the vehicle user's trip, the start and stop times of the vehicle include the start time of the vehicle user's first trip every day and the end time of the vehicle user's last trip every day, and the starting and ending points of the driving trajectory include the starting and ending points of the vehicle user's first trip every day and the starting and ending points of the vehicle user's last trip every day. Method 2: Use an electronic license plate reader to record the driving data of the vehicle corresponding to the recognized electronic license plate, that is, the historical driving data of the vehicle, construct a driving trajectory data cube and a periodic time slice based on the historical driving data of the vehicle, perform hierarchical clustering based on a sliding window on the periodic time slice, and analyze the travel patterns of vehicle users according to the clustering results. Among them, the driving trajectory data cube includes a periodic dimension, a time dimension, and an electronic license plate reader dimension.

[0003] In the patent with the application number 201911329329.1 and the patent name "A Method for Compressing Truck GPS Trajectory Data", a method for compressing truck GPS trajectory data is disclosed, including the following steps: Step 1: Based on the acquired truck GPS trajectory data, perform data cleaning on the truck GPS trajectory data, that is, data screening; Step 2: Based on the truck GPS trajectory data cleaned in Step 1, identify the truck stop points through stop point identification; Step 3: Based on the truck stop point data and GPS trajectory data in Step 2, combined with the road network data, obtain the compression value of the truck trajectory through the trajectory compression method. The specific content of Step 1 includes: Step 1A: Write a cleaning program to set the screening conditions for the truck GPS trajectory data to remove duplicate and invalid GPS data; Step 1B: Clean the truck GPS trajectory data through the cleaning program written in Step 1A; Step 1C: Save the truck GPS trajectory data cleaned in Step 1B in Excel format locally. The specific content of Step 2 includes: Step 2A: Calculate the speed threshold required for identifying suspected stop points in the trajectory data. The calculation formula for the speed threshold vset required for identifying suspected stop points in the trajectory data is as follows: In the formula, n is the number of trajectory points with a speed of 0 km / h in a segment of N trajectory data, and vi-1, vi-2, vi-3, vi+1, Vi+2, vi+3 represent the speed values corresponding to the 3 trajectory points Ti-1, Ti-2, Ti-3 taken upward by serial number and the 3 trajectory points Ti+1, Ti+2, Ti+3 taken downward by serial number for all the points Ti with a speed of 0 km / h in the trajectory data; Step 2B: Based on the truck GPS trajectory data in Step 1 and the speed threshold in Step 2A, compare the instantaneous speed v of the truck's travel with the predefined speed threshold V. If v < V, then take this point as a suspected stop point of the truck; Step 2C: If v > V, then take this point as a driving point; Step 2D: Based on the vehicle type attribute in the GPS trajectory data, determine the normal stop activities of the vehicle. The shortest stop time for the vehicle to perform normal stop activities is the time threshold; Step 2E: Based on the set of suspected stop points of the truck in Step 2B, compare the stop time tn of each truck with the time threshold T in Step 2D. If tn < T, then take this set as the set of stop points; Step 2F: If tn > T, then treat this set of suspected stop points as general trajectory points and perform Step 3; Step 2G: Based on the set of stop points in Step 2E, select the last point in the set of stop points as the stop point in this set, and ignore other trajectory points.Step 3 specifically includes: Step 3A: Select a map open platform with accurate, comprehensive, and open information to download the road network data of China; Step 3B: Based on the platform, download the road network data of China and save it locally; Step 3C: Based on the road network data of China downloaded in Step 3B, use the OSM library in Python to process the road network data, screen out the intersection points between each road in all the road network data, and save them locally; Step 3D: Based on the truck GPS trajectory data in Step 1 and the road network intersection point data in Step 3C, combined with the stop point data in Step 2, compress the truck GPS trajectory data; if there is only one route on the truck trajectory and there is no fork in the road, only record the entry and exit points of the truck trajectory on this road, that is, the intersection points between this road and its adjacent roads; Step 3E: If there is a fork in the road on the truck trajectory, record the trajectory points of the trajectory data at this fork. If there are multiple trajectory points, only retain the trajectory point with the smallest distance from the intersection point; Step 3F: After all the trajectory points of the truck are processed, connect them into a trajectory line according to the stop point data and the serial number attribute of the compressed trajectory record, and finally obtain the compressed trajectory route of the truck.

[0004] A patent with the application number 202080004860.6 and the patent name "Analysis Method, Server and System for Automobile Travel Patterns" discloses an analysis method, server and system for automobile travel patterns, which relates to the field of communication technologies. The method includes: the server obtains the position information of multiple driving trajectories of an automobile; wherein, each driving trajectory includes multiple driving points of the automobile, and the position information of the driving trajectory includes the position information of each driving point; the server determines the trajectory similarity between any two driving trajectories among the multiple driving trajectories according to the position information of each driving trajectory and the trajectory length of each driving trajectory; the server obtains the travel pattern of the automobile according to the trajectory similarity between any two driving trajectories.

[0005] The cleaning of the truck trajectory data in Comparative Document 1 ensures the data quality. Comparative Document 2 sorts out and expresses the travel pattern, but both Comparative Documents 1 and 2 perform interval sampling on the trajectory line according to a certain fixed rule, and only save and use the sampled points to represent the whole trajectory. The compressed trajectory obtained by sampling does not reflect the geometric characteristics of the original trajectory. The reason is that the sampling or pooling standard is static and does not apply a dynamic standard according to the unique characteristics of different trajectories. Summary of the Invention

[0006] The object of the present invention is to solve the deficiencies in the above background technology and provide a compression method for vehicle trajectories that can effectively reflect the geometric characteristics of the original trajectory and express the unique characteristics of different trajectories through dynamic standards.

[0007] To achieve this purpose, the vehicle trajectory compression method designed by the present invention is characterized in that: the compression ratio of the compressed trajectory to the original trajectory is set as n, where n≥1. Taking the starting point of the original trajectory as the starting point of the compressed trajectory, the original trajectory is divided into several original trajectory segments with a length of n (the length unit can be determined according to the actual situation). The several original trajectory segments are compressed into a compressed trajectory composed of several compressed trajectory segments with a length of 1 (the length unit can be determined according to the actual situation), so that the similarity between the compressed trajectory composed of several said compressed trajectory segments and the original trajectory is the largest.

[0008] Further, starting from the starting point of the original trajectory, for each taken original trajectory segment with a length of n, a new coordinate point is added to the compressed trajectory until the original trajectory is completely taken as several said original trajectory segments. Connecting the starting point of the compressed trajectory and the newly added coordinate points in sequence forms the compressed trajectory, and the line segment between two adjacent coordinate points of the compressed trajectory is the compressed trajectory segment.

[0009] Further, the selection criterion for adding a new coordinate point to the compressed trajectory is: within the range of available points for the new coordinate point, a point is selected to maximize the similarity between the compressed trajectory segment after adding this point and the already selected original trajectory segment.

[0010] Further, there are the following three situations for the range of available points for the newly added coordinate point:

[0011] Situation 1: The range of available points for the newly added coordinate point is: all coordinate points between the last coordinate point of the existing compressed trajectory segment and the last coordinate point of the already selected original trajectory segment, excluding the last coordinate point of the existing compressed trajectory segment and including the last coordinate point of the already selected original trajectory segment.

[0012] Situation 2: The range of available points for the newly added coordinate point is: all coordinate points between the last coordinate point of the existing compressed trajectory segment and the last first-order coordinate point of the already selected original trajectory segment, excluding the last coordinate point of the existing compressed trajectory segment and including the last first-order coordinate point of the already selected original trajectory segment. The first-order coordinate points of the original trajectory segment include the mean points of the coordinate points of multiple adjacent two original trajectory segments.

[0013] Situation 3: The range of available points for the newly added coordinate point is: all coordinate points between the last coordinate point of the existing compressed trajectory segment and the last N-order coordinate point of the already selected original trajectory segment, where N≥2, excluding the last coordinate point of the existing compressed trajectory segment and including the last N-order coordinate point of the already selected original trajectory segment. The N-order coordinate points of the original trajectory segment include the mean points of the (N - 1)-order coordinate points of multiple adjacent two original trajectory segments.

[0014] Further, the calculation method for the N - order coordinate points of the original trajectory segment is as follows: calculate the first - order coordinate points of the original trajectory segment to obtain a set of first - order coordinate points of the original trajectory segment, and based on this set of first - order coordinate points of the original trajectory segment, calculate a set of multi - order coordinate points of the second order and above of the original trajectory segment;

[0015] Specifically, the calculation method for the first - order coordinate points of the original trajectory segment is: calculate the mean points of the coordinate points of two adjacent original trajectory segments. The set of first - order coordinate points of the original trajectory segment includes multiple mean points of the coordinate points of two adjacent original trajectory segments.

[0016] Further, the method for judging the similarity between the compressed trajectory segment after adding this point and the selected original trajectory segment is: calculate the distance between the compressed trajectory segment after adding this point and the selected original trajectory segment. The smaller the distance, the greater the similarity between the compressed trajectory segment after adding this point and the selected original trajectory segment.

[0017] Preferably, the minimum distance between the compressed trajectory segment after adding this point and the selected original trajectory segment is the DTW distance. DTW distance: Dynamic Time Warping distance.

[0018] Specifically, the compression method of the vehicle trajectory includes the following steps:

[0019] Step 1: Set the compression ratio of the compressed trajectory to the original trajectory as n, where n≥1. Use the starting point of the original trajectory as the starting point of the compressed trajectory, and divide the original trajectory into several original trajectory segments with a length of n;

[0020] Step 2: Starting from the starting point of the original trajectory, for each original trajectory segment with a length of n taken, add a new coordinate point to the compressed trajectory, so that the distance between the compressed trajectory segment after adding the new coordinate point and the taken original trajectory segment is the DTW distance;

[0021] Step 3: Repeat Step 2 until the original trajectory is completely taken as several original trajectory segments, and connect the starting point of the compressed trajectory and each newly added coordinate point in sequence to form the compressed trajectory.

[0022] The beneficial effects of the present invention are as follows: The compression ratio of the given compression trajectory to the original trajectory is n:1, where n≥1, denoted as n. Initially, the starting point of the compression trajectory is the same as that of the original trajectory. The original trajectory only takes the first segment with a length of n, and the length of the compression trajectory is 1. After that, the original trajectory takes the next segment with a length of n, and the compression trajectory adds a coordinate point until the original trajectory is completely taken. At this time, the corresponding compression trajectory is the calculation result. In the present invention, the key is to select the next coordinate point of the compression path according to a dynamic selection criterion. For this purpose, this coordinate point can be directly selected from the original coordinate points, or from the points obtained by performing a certain processing on the original coordinate points. For example, the original coordinate points are set in groups of k (k≥2), and the center point of each group is calculated, and the next coordinate point added to the compression path is selected from these center points. The selection criterion is that among all the optional coordinate points, the selected one makes the compression path after adding it have the greatest similarity to the currently existing original trajectory. Any method for measuring the distance between two time series of unequal lengths can be used to evaluate the similarity between the compression path and the original trajectory. The smallest distance is considered to have the greatest similarity. For example, the DTW distance between the compression trajectory and the original trajectory is the smallest. During the generation process of the compression trajectory, the compression trajectory segment and the original trajectory segment grow synchronously. When a segment (the segment length is determined by the compression ratio) is added to the original trajectory segment, a point is added to the compression trajectory segment. The point added to the compression trajectory segment is selected from all the optional points. The distance between the compression trajectory segment and the original trajectory segment after adding this point is calculated, and the point that can minimize the distance between the two trajectories is selected as the optimal point and added to the compression trajectory segment, so as to always maintain the highest similarity between the existing compression trajectory segment and the original trajectory segment at each step. In summary, the present invention selects the coordinate points of the compression trajectory according to a dynamic measurement criterion, maximally ensuring the similarity between the compression trajectory and the original trajectory, enabling the compression trajectory to always reflect the uniqueness of the original trajectory. The compression trajectory has a high degree of restoration to the original trajectory. On the basis of ensuring the effective compression and storage of the original trajectory, the travel pattern of the vehicle can be accurately judged according to the compression trajectory, laying a foundation for the actual application of big data in the later stage, and having good market application value. Description of the Drawings

[0023] Figure 1 It is a schematic diagram of the value-taking of multi-order coordinate points of the original trajectory in the present invention. Detailed Embodiment

[0024] The following further elaborates on the present invention in detail in conjunction with the drawings and specific embodiments. The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. A method for compressing a vehicle trajectory includes the following steps:

[0025] Step 1: Set the compression ratio of the compression trajectory to the original trajectory as n, where n≥1. Take the starting point of the original trajectory as the starting point of the compression trajectory, and divide the original trajectory into several original trajectory segments with a length of n (the length unit can be determined according to the actual situation).

[0026] Step 2: Starting from the starting point of the original trajectory, for each original trajectory segment with a length of n taken, a new coordinate point is added to the compression trajectory. The selection criterion for adding a new coordinate point to the compression trajectory is as follows: within the range of available points for the new coordinate point, select a point that maximizes the similarity between the compression trajectory segment after adding this point and the selected original trajectory segment. The method for judging the similarity between the compression trajectory segment after adding this point and the selected original trajectory segment is: calculate the distance between the compression trajectory segment after adding this point and the selected original trajectory segment. The smaller the distance, the greater the similarity between the compression trajectory segment after adding this point and the selected original trajectory segment. In the present invention, the method for calculating the maximum similarity between the compression trajectory and the selected original trajectory is to calculate the DTW distance (DTW distance: Dynamic Time Warping distance) between the two trajectories, that is, the distance between the compression trajectory segment after adding the new coordinate point and the taken original trajectory segment is the DTW distance.

[0027] Step 3: Repeat Step 2 until the original trajectory is completely taken as several original trajectory segments, and connect the starting point of the compression trajectory and each newly added coordinate point in sequence to form the compression trajectory.

[0028] In the above solution, there are the following three situations for the range of available points for adding a new coordinate point:

[0029] Situation 1: All coordinate points between the last coordinate point of the existing compression trajectory segment and the last coordinate point of the selected original trajectory segment, excluding the last coordinate point of the existing compression trajectory segment and including the last coordinate point of the selected original trajectory segment.

[0030] Situation 2: All coordinate points between the last coordinate point of the existing compression trajectory segment and the last first-order coordinate point of the selected original trajectory segment, excluding the last coordinate point of the existing compression trajectory segment and including the last first-order coordinate point of the selected original trajectory segment. The first-order coordinate points of the original trajectory segment include the mean points of the coordinate points of multiple adjacent two original trajectory segments.

[0031] Situation 3: All coordinate points between the last coordinate point of the existing compression trajectory segment and the last N-order coordinate point of the selected original trajectory segment, where N≥2, excluding the last coordinate point of the existing compression trajectory segment and including the last N-order coordinate point of the selected original trajectory segment. The N-order coordinate points of the original trajectory segment include the mean points of the (N - 1)-order coordinate points of multiple adjacent two original trajectory segments.

[0032] In Case 3, the calculation method for the Nth-order coordinate points of the original trajectory segment is as follows: Calculate the first-order coordinate points of the original trajectory segment to obtain a set of first-order coordinate points of the original trajectory segment. Based on the set of first-order coordinate points of the original trajectory segment, calculate a set of second-order and higher-order coordinate points of the original trajectory segment;

[0033] Among them, the calculation method for the first-order coordinate points of the original trajectory segment is: Calculate the mean points of the coordinate points of two adjacent original trajectory segments. The set of first-order coordinate points of the original trajectory segment includes multiple mean points of the coordinate points of two adjacent original trajectory segments.

[0034] The following specifically describes the vehicle trajectory compression method designed by the present invention through two embodiments:

[0035] Embodiment 1:

[0036] Prepare two sequences A and B:

[0037] Compression ratio: For every n coordinate points in the original trajectory, there is only one coordinate point in the compressed path.

[0038] 1. Initialization: Put the first n coordinate points of the original trajectory into A, and put the first coordinate point of the original trajectory into B. Step serial number step_index = 1

[0039] A is used to store the original trajectory. At the beginning, put the n coordinate points of the original trajectory, and then add n points in each step until all the complete original trajectory is put in; B is used to store the compressed trajectory as the result. Initially, put the initial point of the original trajectory, and then put one coordinate point in each subsequent step.

[0040] 2. Select the next coordinate point to be added to B:

[0041] The serial numbers of the coordinate points placed in A and B remain the same as those in the original trajectory. Since A continuously puts in all the coordinate points of the original trajectory, the serial numbers in A are continuous; the coordinate points placed in B are selected from the original trajectory, so the serial numbers in B are non - continuous.

[0042] The serial number of the last coordinate point in A currently is A_index_i = step_index * n; the serial number of the last coordinate point in B currently is B_index_i.

[0043] Select the next coordinate point B_index_i+1 of B from the coordinate points within the serial number range of (B_index_i, A_index_i] of the original trajectory. Among them, the selection range of B_index_i+1 is not limited to the newly added n coordinate points of A, but from the last coordinate point of B (excluding this point) to the last one (including) of the newly added n coordinate points of A, and all are within the selection range. There is no requirement that the points added to B must be one of the newly added n coordinate points of A. Due to the absence of such a rigid correspondence, it will not cause unnecessary data distortion.

[0044] 3. Update A and B: Add the selected B_index_i+1 in step 2 to B, add the next n coordinate points to A, and increment Step_index by 1.

[0045] 4. Repeat steps 2 and 3 until all original trajectories are placed in A.

[0046] 5. Return B as the compressed trajectory.

[0047] Embodiment 2:

[0048] The difference between Embodiment 2 and Embodiment 1 is that when selecting the next point to be added in B, it is not selected from the original coordinate points of the original trajectory, but from the first-order (second-order or better) combined points of the original coordinate points. As Figure 1 shown, the circles at the bottom layer represent the coordinate points of the original trajectory, the circles at the second layer represent the first-order coordinate points of the original trajectory (the mean points of two adjacent original trajectory coordinate points), and the circles at the third layer represent the second-order coordinate points (the mean points of two adjacent first-order coordinate points of the original trajectory).

[0049] When it is set to take the first-order coordinate points, when selecting the next point to be added in B, it is selected from all the first-order coordinate points; when it is set to take the second-order coordinate points, it is selected from all the second-order combined points; the higher the order, the fewer the alternative points, the faster the algorithm executes, but the higher the degree of data distortion.

[0050] Here, it should be noted that the description of the above technical solution is exemplary. This specification can be embodied in different forms and should not be construed as limited to the technical solutions set forth herein. On the contrary, providing these descriptions will make the disclosure of the present invention thorough and complete, and will fully convey the scope disclosed in this specification to those skilled in the art. In addition, the technical solutions of the present invention are only defined by the scope of the claims. The shapes, sizes, ratios, angles, and numbers disclosed for describing various aspects of this specification and the claims are merely examples. Therefore, this specification and the claims are not limited to the details shown. In the following description, when the detailed description of related known functions or configurations is determined to unnecessarily obscure the focus of this specification and the claims, the detailed description will be omitted. When using "including", "having", and "comprising" described in this specification, there may also be another part or other parts, and the terms used can generally be singular but may also represent plural forms.

[0051] In the present invention, the compression ratio of the given compression trajectory to the original trajectory is n:1, denoted as n. Initially, the starting point of the compression trajectory is the same as that of the original trajectory. The original trajectory only takes the first segment with a length of n, and the length of the compression trajectory is 1. Thereafter, the original trajectory takes the next segment with a length of n, and the compression trajectory adds a coordinate point until the original trajectory is completely taken. At this time, the corresponding compression trajectory is the calculation result. In the present invention, the key is to select the next coordinate point of the compression path according to a dynamic selection criterion. For this purpose, this coordinate point can be directly selected from the original coordinate points, or from the points obtained by performing a certain processing on the original coordinate points. For example, the original coordinate points are set in groups of k (k≥2), the center point of each group is calculated, and the next coordinate point added to the compression path is selected from these center points. The selection criterion is that among all the optional coordinate points, the selected one makes the compression path after adding it have the highest similarity to the currently existing original trajectory. Any method for measuring the distance between two unequal-length time series can be used to evaluate the similarity between the compression path and the original trajectory. The smallest distance is considered to have the highest similarity. For example, the DTW distance between the compression trajectory and the original trajectory is the smallest. During the generation process of the compression trajectory, the compression trajectory segment and the original trajectory segment grow synchronously. When a segment (the segment length is determined by the compression ratio) is added to the original trajectory segment, a point is added to the compression trajectory segment. The point added to the compression trajectory segment is selected from all the optional points. The distance between the compression trajectory segment and the original trajectory segment after adding this point is calculated, and the point that can make the distance between the two trajectories the smallest is taken as the optimal point and added to the compression trajectory segment, so as to always maintain the highest similarity between the existing compression trajectory segment and the original trajectory segment at each step. To sum up, the present invention selects the coordinate points of the compression trajectory according to a dynamic measurement criterion, maximally ensures the similarity between the compression trajectory and the original trajectory, enables the compression trajectory to always reflect the uniqueness of the original trajectory, has a high degree of restoration of the compression trajectory to the original trajectory, and on the basis of ensuring the effective compression storage of the original trajectory, can accurately judge the travel rules of the vehicle according to the compression trajectory, laying a foundation for the actual application of big data in the later stage and having good market application value.

[0052] Finally, it should be pointed out that the above embodiments are only relatively representative examples of the present invention. Obviously, the present invention is not limited to the above embodiments and there can be many variations. Any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention should be considered to fall within the protection scope of the present invention.

Claims

1. A method for compressing vehicle trajectories, characterized in that: Set the compression ratio of the compression trajectory to the original trajectory as n, where n ≥ 1. Take the starting point of the original trajectory as the starting point of the compression trajectory. Divide the original trajectory into several original trajectory segments of length n, and compress the several original trajectory segments into several compression trajectory segments of length 1, so that the similarity between the compression trajectory composed of the several compression trajectory segments and the original trajectory is the largest. Starting from the starting point of the original trajectory, for each original trajectory segment of length n taken, a new coordinate point is added to the compression trajectory until the original trajectory is completely taken as several original trajectory segments. Connect the starting point of the compression trajectory and the newly added coordinate points in sequence to form the compression trajectory. The line segment between two adjacent coordinate points of the compression trajectory is the compression trajectory segment. The selection criterion for adding a new coordinate point to the compression trajectory is: within the range of available points for the new coordinate point, select a point to maximize the similarity between the compression trajectory segment after adding this point and the selected original trajectory segment.

2. The method for compressing a vehicle trajectory according to claim 1, characterized in that: The range of available points for the newly added coordinate point is: all coordinate points between the last coordinate point of the existing compression trajectory segment and the last coordinate point of the selected original trajectory segment, excluding the last coordinate point of the existing compression trajectory segment and including the last coordinate point of the selected original trajectory segment.

3. The method for compressing a vehicle trajectory according to claim 1, wherein: The range of available points for the newly added coordinate point is: all coordinate points between the last coordinate point of the existing compression trajectory segment and the last first-order coordinate point of the selected original trajectory segment, excluding the last coordinate point of the existing compression trajectory segment and including the last first-order coordinate point of the selected original trajectory segment. The first-order coordinate points of the original trajectory segment include the mean points of the coordinate points of multiple adjacent two original trajectory segments.

4. The method for compressing a vehicle trajectory according to claim 3, wherein: The range of available points for the newly added coordinate point is: all coordinate points between the last coordinate point of the existing compression trajectory segment and the last N-order coordinate point of the selected original trajectory segment, where N ≥ 2, excluding the last coordinate point of the existing compression trajectory segment and including the last N-order coordinate point of the selected original trajectory segment. The N-order coordinate points of the original trajectory segment include the mean points of the (N - 1)-order coordinate points of multiple adjacent two original trajectory segments.

5. The method for compressing a vehicle trajectory according to claim 4, wherein: The calculation method for the N-order coordinate points of the original trajectory segment is: calculate the first-order coordinate points of the original trajectory segment to obtain the set of first-order coordinate points of the original trajectory segment. Based on the set of first-order coordinate points of the original trajectory segment, calculate the set of second-order and higher-order coordinate points of the original trajectory segment. The calculation method for the first-order coordinate points of the original trajectory segment is: calculate the mean points of the coordinate points of adjacent two original trajectory segments. The set of first-order coordinate points of the original trajectory segment includes the mean points of the coordinate points of multiple adjacent two original trajectory segments.

6. The method for compressing a vehicle trajectory according to claim 2 or 3 or 4, characterized in that: The judgment method for the similarity between the compression trajectory segment after adding this point and the selected original trajectory segment is: calculate the distance between the compression trajectory segment after adding this point and the selected original trajectory segment. The smaller the distance, the greater the similarity between the compression trajectory segment after adding this point and the selected original trajectory segment.

7. The method for compressing a vehicle trajectory according to claim 6, characterized in that: The minimum distance between the compressed trajectory segment after adding this point and the selected original trajectory segment is the DTW distance.

8. The method for compressing a vehicle trajectory according to claim 7, characterized in that: It includes the following steps: Step 1: Set the compression ratio of the compressed trajectory to the original trajectory as n, where n ≥ 1. Taking the starting point of the original trajectory as the starting point of the compressed trajectory, divide the original trajectory into several original trajectory segments with a length of n. Step 2: Starting from the starting point of the original trajectory, for each original trajectory segment with a length of n taken, a new coordinate point is added to the compressed trajectory, so that the distance between the compressed trajectory segment after adding the new coordinate point and the taken original trajectory segment is the DTW distance. Step 3: Repeat Step 2 until the original trajectory is completely taken as several original trajectory segments, and connect the starting point of the compressed trajectory and each newly added coordinate point in sequence to form the compressed trajectory.

Images