A vehicle multi-section trajectory matching method and system based on radar detection

Abstract

The present invention provides a vehicle multi-section trajectory matching method and system based on radar detection, belonging to the field of intelligent transportation. The method first obtains the trajectory data of the intersection detected by the radar, and obtains the global trajectory pool of the vehicle after preprocessing; an independent entrance trajectory pool and exit trajectory pool are set at each intersection, and the trajectory data is distributed to the trajectory pool, and then the trajectories in the entrance trajectory pool and the exit trajectory of the same intersection are paired, and according to the calculated trajectory distance of the trajectory pair, the trajectory pair with a trajectory distance less than a threshold is taken out, and after successful matching, the trajectory of the current vehicle at the current intersection is generated, and the successfully matched trajectory pairs are cleaned up in time. The present invention only needs to calibrate the central longitude and latitude of each intersection, does not need to calibrate the lane range, the allowed driving direction of the lane, and the upstream and downstream adjacency relationship of the section, does not require overlapping coverage of different radars, is simple to implement, has loose data quality requirements, is flexible to implement, has a wide range of applications, and is highly practical.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation, and specifically relates to a vehicle multi-section trajectory matching method and system based on radar detection. Background Art

[0002] When tracking and navigating a vehicle, it is necessary to accurately grasp the real-time position and complete trajectory of the vehicle. Vehicles are generally tracked by trajectory matching. Multi-section trajectory matching refers to the problem of finding trajectory records from the same traffic unit from trajectory detection records under different radars when a millimeter-wave radar only covers a section with adjacent intersections as endpoints.

[0003] In the prior art, there are a variety of road section trajectory matching methods based on radar detection. For example, application number CN202111026263.6 discloses a method for matching and fusion of multi-radar detection trajectories at intersections. By matching the actual trajectory detected by the radar with the preset reference trajectory, multiple trajectory fragments of the same vehicle detected within the same radar and between different radars are matched and fused in real time, thereby outputting the complete driving trajectory of the same vehicle passing through the intersection. However, when matching the above trajectories, lane-level radar data is required, which has high requirements for data quality, and the matching effect is heavily dependent on the data. Application number CN202310094737.3 discloses a highway vehicle trajectory splicing method based on radar data, which obtains vehicle trajectory data of multiple radars in real time and performs preprocessing and track division, and coordinates the position of road vehicles within the field of view of each radar. After the coordinate system is unified, the vehicle trajectory data processed by two adjacent radars is traversed, and the unified trajectory points of the two trajectories are determined according to the matching conditions, and the trajectory is spliced. However, the above method is applicable to vehicles that are divided into lanes on highways, and its universality is not strong. Existing methods generally have the problem of being heavily dependent on data. When radars are sparsely installed and the coverage range does not overlap, the complete vehicle trajectory cannot be obtained and the matching effect is poor. Summary of the invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, the present invention aims to provide a vehicle multi-section trajectory matching method and system based on radar detection. By setting an entry trajectory pool and an exit trajectory pool, and setting corresponding thresholds for trajectory matching, when the radar coverage is small and part of it cannot overlap, a relatively accurate trajectory matching can still be obtained to obtain a complete vehicle route, thereby solving the problem of maintaining the vehicle ID between different radars when the radar coverage only includes road sections and the overlapping area is small.

[0005] In order to achieve the above purpose, the embodiment of the present invention adopts the following technical solution:

[0006] In a first aspect, an embodiment of the present invention provides a vehicle multi-segment trajectory matching method based on radar detection, the method comprising the following steps:

[0007] Step S1, obtaining the track data of the intersection detected by the radar, and when the original track data detected by the radar is not read after a predetermined time, it is determined that the track data detected by the radar for the vehicle has all arrived;

[0008] Step S2, establish a global trajectory pool S all ; Preprocess all the trajectory data that have arrived, obtain the trajectory id, and set the matching identification field corresponding to the trajectory id; and put all the trajectory data, the corresponding trajectory id and the matching identification field into S all Among them, S all Includes all the trajectories tracked and reached at all intersections;

[0009] Step S3: two independent trajectory pools are set at each intersection, namely, the entrance trajectory pool S in and exit trajectory pool S out , S all The trajectories in the S are assigned to the corresponding intersections according to the predetermined logic. in or S out middle;

[0010] Step S4, traverse each intersection in turn, and compare the S in and S out Pair the trajectories in the image with each other, calculate the corresponding trajectory distances, and sort all the calculated trajectory distances of the same intersection in ascending order, and take out the trajectory distances less than the threshold σ 2 The trajectory pair;

[0011] Step S5, identifying the trajectory data in the extracted trajectory pair, and if the trajectory data has not been matched before, determining that the trajectory pair is matched successfully;

[0012] Step S6, the matched trajectory is aligned, and the out Set start_matched = true to match the trajectory extracted from S in The trajectory taken from the trajectories is set to end_matched = true matching flag; the successful matching result is recorded as output, and the successful matching trajectory pairs are respectively taken from the S of the intersection in and S out Clean it out.

[0013] As a preferred embodiment of the present invention, the predetermined logic in step S3 includes the following steps:

[0014] Step S31, if the matching flag of the trajectory is start_matched = true, skip the current trajectory; otherwise, execute the following steps: take the intersection center as the intersection coordinate value, calculate the distance of each intersection from the starting point of the trajectory data corresponding to the current trajectory id and sort them, if the distance between the starting point and the nearest intersection is less than the threshold σ 1 , then the trajectory is assigned to the intersection S closest to the starting point out , otherwise skip;

[0015] Step S32, if the matching flag of the trajectory end_matched = true, skip the current trajectory; otherwise, execute the following steps: calculate the distance between each intersection and the end point of the trajectory data corresponding to the current trajectory id and sort them. If the distance between the end point and the nearest intersection is less than the threshold σ 1 , then the trajectory is assigned to the intersection S closest to the end point in , otherwise skip;

[0016] Step S33: If a trajectory is not assigned to any intersection, in or S out , it is considered that the vehicle has left the road network in the middle of a certain road section, and no operation is performed on the vehicle, waiting until step S6 to clear the track.

[0017] As a preferred embodiment of the present invention, when calculating the corresponding trajectory distance in step S4, both the time and space information of the trajectory are considered, and the specific process is as follows:

[0018] Step S41, given two trajectories in is the i-th trajectory point of the j-th trajectory, are the longitude, latitude and timestamp of the i-th track point of the j-th track respectively;

[0019] Step S42, determine the two trajectories T 1 and T 2 Are the radar ids in the trajectory ids the same? If they are the same, let T 1 With T 2 If the trajectory distance between them is positive infinity, the calculation ends; if not, execute step S43;

[0020] Step S43: 1 In T 2 Extrapolate on timestamps where time is greater than Less than The fragments are filled using linear interpolation, less than The fragment is extended by linear interpolation of the first two trajectory points, which is greater than The segment is extended by linear interpolation of the two ending trajectory points to obtain the extended trajectory T 1 ′;

[0021] Step S44, greater than The epitaxial fragment is rotated 45°, 90°, -45° and -90° to obtain T′ 1a ,T′ 1b ,T′ 1c ,T′ 1d Four extended trajectories simulating turns;

[0022] Step S45, calculate T' 1 ,T′ 1a ,T′ 1b ,T′ 1c ,T′ 1d Respectively with T 2 The smallest value is taken as T 1 and T 2 The track distance between them.

[0023] As a preferred embodiment of the present invention, when acquiring radar detection data, every τ input s reads the trajectory input generated during the monitored time period, τ input The time interval for reading data.

[0024] As a preferred embodiment of the present invention, the fields contained in the original trajectory data in step S1 include: radar id, vehicle id, timestamp, longitude, latitude and speed.

[0025] As a preferred embodiment of the present invention, the trajectory ID includes a radar ID, a vehicle ID and a start timestamp.

[0026] As a preferred embodiment of the present invention, step S5 further includes: at a preset time point, traversing the global trajectory pool S all If there is a trajectory that has not been matched for a long time, it will be removed from the global trajectory pool S all Clean it out.

[0027] In a second aspect, an embodiment of the present invention further provides a vehicle multi-section trajectory matching system based on radar detection, the system comprising: a data acquisition module, a global trajectory pool establishment module, an independent trajectory pool setting module, a trajectory allocation module, a trajectory distance calculation module, a matching identification module, and a post-processing module; wherein,

[0028] The data acquisition module is used to acquire the trajectory data of the intersection detected by the radar. When the original trajectory data detected by the radar is not read after a predetermined period of time, it is determined that the trajectory data detected by the radar for the vehicle has all arrived;

[0029] The global trajectory pool establishment module is used to establish a global trajectory pool S all ; Preprocess all the trajectory data that have arrived, obtain the trajectory id, and set the matching identification field corresponding to the trajectory id; and put all the trajectory data, the corresponding trajectory id and the matching identification field into S all Among them, S all Includes all the trajectories tracked and reached at all intersections;

[0030] The independent track pool setting module is used to set two independent track pools at each intersection, namely the entrance track pool S in and exit trajectory pool S out ;

[0031] The trajectory allocation module is used to allocate S all The trajectories in the S are assigned to the corresponding intersections according to the predetermined logic. in or S out middle;

[0032] The trajectory distance calculation module is used to traverse each intersection in turn and calculate the S in and S out Pair the trajectories in the image with each other, calculate the corresponding trajectory distances, and sort all the calculated trajectory distances of the same intersection in ascending order, and take out the trajectory distances less than the threshold σ 2 The trajectory pair;

[0033] The matching identification module is used to identify the trajectory data in the extracted trajectory pair, and if the trajectory data has not been matched before, it is determined that the trajectory pair is matched successfully;

[0034] The post-processing module is used to center the successfully matched trajectory from S out Set start_matched = true to match the trajectory extracted from S in The trajectory taken from the trajectories is set to end_matched = true matching flag, the successful matching result is recorded as output, and the successful matching trajectory pairs are respectively taken from the intersection S in and S out Clean it out.

[0035] As a preferred embodiment of the present invention, the trajectory distance calculation module considers both the time and space information of the trajectory when calculating the corresponding trajectory distance.

[0036] The technical solution provided by the embodiment of the present invention has the following beneficial effects:

[0037] The vehicle multi-section trajectory matching method and system based on radar detection provided by the embodiment of the present invention, by setting an entry pool and an exit pool at the intersection, calculating the distance of the trajectory pairs in the entry pool and the exit pool, and performing real-time matching on the same vehicle detected by different radars, so as to achieve the vehicle driving from one section to another, and the ID under different radars remains unchanged, and obtain the complete driving trajectory of the vehicle. The present invention is simple to implement, has loose data quality requirements, and is flexible to implement. It only needs to calibrate the central longitude and latitude of each intersection, and does not need to calibrate other information such as the lane range, the allowed driving direction of the lane, and the upstream and downstream adjacency relationship of the section; it does not require that the coverage ranges of different radars overlap, and the requirements for data quality are not harsh; when implemented, it can be deployed for a road network containing hundreds of intersections or for several adjacent or non-adjacent intersections. In places where the radar coverage ranges intersect but there are no intersections, intersections can be manually added to complete the deployment. It has high flexibility, a wide range of applications, and strong practical capabilities.

[0038] Of course, it is not necessary to achieve all of the advantages described above at the same time to implement any product or method of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 is a flow chart of a vehicle multi-section trajectory matching method based on radar detection in an embodiment of the present invention;

[0041] Figure 2 It is a schematic diagram of a road intersection scene of a specific application example of the present invention. DETAILED DESCRIPTION

[0042] The technical solutions in the embodiments of the present invention will be described clearly and completely 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. The components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations. It should be noted that the embodiments of the present invention and the features in the embodiments can also be combined with each other without conflict.

[0043] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In the description of the present invention, the terms "first", "second", "third", "fourth", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0044] Based on the problem that vehicle trajectory matching cannot be maintained when the radar coverage is small and there is a mismatch, an embodiment of the present invention provides a vehicle multi-section trajectory matching method based on radar detection. By setting an independent entrance trajectory pool and exit trajectory pool at each intersection, the distance between the trajectory pairs in the entrance trajectory pool and the exit trajectory pool is calculated, and the same vehicle detected by different radars is matched in real time, so that the vehicle can travel from one section to another, and the ID under different radars remains unchanged, thereby obtaining the complete driving trajectory of the vehicle.

[0045] like Figure 1 As shown, the vehicle multi-segment trajectory matching method based on radar detection provided by an embodiment of the present invention includes the following steps:

[0046] Step S1, obtaining the trajectory data of the intersection detected by the radar. When the trajectory input of the vehicle generated by the radar is not read after a predetermined period of time, it is determined that the trajectory data detected by the radar for the vehicle has all arrived.

[0047] In this step, when accessing and acquiring radar detection data, every τ input s reads the trajectory input generated during the monitored time period, τ input The time interval for data reading, measured in seconds.

[0048] The original trajectory data includes fields: radar id, vehicle id, timestamp, longitude, latitude, speed, etc.

[0049] Step S2, establish a global trajectory pool S all ; Preprocess all the trajectory data that have arrived, obtain the trajectory id, and set the matching identification field corresponding to the trajectory id; and put the trajectory data, the corresponding trajectory id and the matching identification field into the global trajectory pool S all Among them, S all The trajectories of all intersections tracked and reached are included; the matching identification field is divided into start_matched and end_matched, and the values ​​are true or false respectively.

[0050] In this step, the preprocessing of the trajectory data includes: reading the trajectory data of all intersections, generating a trajectory id according to the radar id, vehicle id and start timestamp carried by each trajectory data, and the format of the trajectory id is "radar id + vehicle id + start timestamp of the trajectory"; at the same time, setting the matching identification field corresponding to the trajectory id.

[0051] Step S3: two independent trajectory pools are set at each intersection, namely, the entrance trajectory pool S in and exit trajectory pool S out , the global trajectory pool S all The trajectories in are assigned to the entry trajectory pool S of the corresponding intersection according to the predetermined logic. in Or exit trajectory pool S out middle.

[0052] Preferably, the predetermined logical steps for performing trajectory allocation in this step are as follows:

[0053] Step S31, if the matching flag of the trajectory start_matched = true, skip it, otherwise proceed to the following steps: take the intersection center as the intersection coordinate value, calculate the distance of each intersection from the starting point of the trajectory data corresponding to the current trajectory id and sort them, if the distance between the starting point and the nearest intersection is less than the threshold σ 1 , then the trajectory is assigned to the exit trajectory pool S of the intersection closest to the starting point out , otherwise skip;

[0054] Step S32, if the matching flag of the trajectory end_matched = true, skip it, otherwise proceed to the following steps: calculate the distance between each intersection and the end point of the trajectory data corresponding to the current trajectory id and sort them, if the distance between the end point and the nearest intersection is less than the threshold σ 1 , then the trajectory is assigned to the entrance trajectory pool S of the intersection closest to the end point in , otherwise skip;

[0055] Step S33, if a trajectory is not assigned to the entrance pool and exit pool of any intersection, it is considered that the vehicle has exited the road network in the middle of a certain section, and no operation is performed on it, and the trajectory is cleared until step S6.

[0056] Step S4, traverse each intersection in turn, and collect the entry trajectory pool S of the same intersection in and exit trajectory pool S out Pair the trajectories in the image with each other, calculate the corresponding trajectory distances, and sort all the calculated trajectory distances of the same intersection in ascending order, and take out the trajectory distances less than the threshold σ 2 The trajectory pair.

[0057] In this step, the time and space information of the trajectory are considered simultaneously when calculating the corresponding trajectory distance. The specific process is as follows:

[0058] Step S41, given two trajectories in is the i-th trajectory point of the j-th trajectory, are the longitude, latitude and timestamp of the i-th trajectory point respectively;

[0059] Step S42, determine the two trajectories T 1 and T 2 Are the radar ids in the trajectory ids the same? If they are the same, let T 1 With T 2 If the trajectory distance between them is positive infinity, the calculation ends; if not, execute step S43;

[0060] Step S43: T 1 In T 2 Extrapolate on timestamps where time is greater than Less than The fragments are filled using linear interpolation, less than The fragment is extended by linear interpolation of the first two trajectory points, which is greater than The segment is extended by linear interpolation of the two ending trajectory points to obtain the extended trajectory T 1 ′;

[0061] Step S44, greater than The epitaxial fragment of T is rotated 45°, 90°, -45° and -90° to obtain 1 ′ a ,T 1 ′ b ,T 1 ′ c ,T 1 ′ d Four extended trajectories simulating turns;

[0062] Step S45, calculate T 1 ′ ,T 1 ′ a ,T 1 ′ b ,T 1 ′ c ,T 1 ′ dRespectively with T 2 The smallest value is taken as T 1 and T 2 The track distance between them.

[0063] Step S5, identifying the trajectory data in the extracted trajectory pair, and if the trajectory data has not been matched before, it is determined that the trajectory pair is matched successfully.

[0064] Step S6: The matched trajectory is collected from the exit trajectory pool S out The trajectory taken from the entry track pool S is set to start_matched = true matching flag. in The trajectory taken from the trajectories is set to end_matched = true matching flag, the successful matching result is recorded as output, and the successful matching trajectory pairs are respectively taken from the intersection S in and S out Clean it out.

[0065] Furthermore, this step may also include: at a preset time point, traversing the global trajectory pool S all If there is a trajectory that is not matched within the predetermined time, the trajectory is removed from the global trajectory pool S all Clean it out.

[0066] Based on the above method, an embodiment of the present invention also provides a vehicle multi-section trajectory matching system based on radar detection, the system comprising: a data acquisition module, a global trajectory pool establishment module, an independent trajectory pool setting module, a trajectory allocation module, a trajectory distance calculation module, a matching identification module, and a post-processing module; wherein,

[0067] The data acquisition module is used to acquire the trajectory data of the intersection detected by the radar. When the original trajectory data detected by the radar is not read after a predetermined period of time, it is determined that the trajectory data detected by the radar for the vehicle has all arrived;

[0068] The global trajectory pool establishment module is used to establish a global trajectory pool S all ; Preprocess all the trajectory data that have arrived, obtain the trajectory id, and set the matching identification field corresponding to the trajectory id; and put all the trajectory data, the corresponding trajectory id and the matching identification field into S all Among them, S all The trajectories of all intersections tracked and reached are included; the matching identification field is divided into start_matched and end_matched, and the values ​​are true or false respectively;

[0069] The independent track pool setting module is used to set two independent track pools at each intersection, namely the entrance track pool Sin and exit trajectory pool S out ;

[0070] The trajectory allocation module is used to allocate S all The trajectories in the S are assigned to the corresponding intersections according to the predetermined logic. in or S out middle;

[0071] The trajectory distance calculation module is used to traverse each intersection in turn and calculate the S in and S out Pair the trajectories in the image with each other, calculate the corresponding trajectory distances, and sort all the calculated trajectory distances of the same intersection in ascending order, and take out the trajectory distances less than the threshold σ 2 The trajectory pair;

[0072] The matching identification module is used to identify the trajectory data in the extracted trajectory pair, and if the trajectory data has not been matched before, it is determined that the trajectory pair is matched successfully;

[0073] The post-processing module is used to center the successfully matched trajectory from S out Set start_matched = true to match the trajectory extracted from S in The trajectory taken from the trajectories is set to end_matched = true matching flag, the successful matching result is recorded as output, and the successful matching trajectory pairs are respectively taken from the intersection S in and S out Clean it out.

[0074] It should be noted that the vehicle multi-segment trajectory matching system based on radar detection provided in the embodiment of the present invention corresponds to the vehicle multi-segment trajectory matching method based on radar detection, and the description of the method is also applicable to the system.

[0075] The present invention is further described in detail below through a specific example. Figure 2 As shown, an intersection is set with three radar detection systems, and the longitude and latitude of the center of the intersection are calibrated to (115.912152, 39.068015).

[0076] Let τ input =2, read the trajectory data of the intersection detected by the radar every 2s within the preset time period. If the original trajectory data detected by the radar is not read after the preset time, it is determined that the trajectory data detected by the radar for the vehicle has been fully reached, and all the data of the trajectory is obtained. The original trajectory data obtained at this time includes several data points in each trajectory, and each data point includes the following fields: radar id, vehicle id, timestamp, longitude, latitude and speed, etc.

[0077] Establish a global trajectory pool S all . Preprocess all the trajectory data that have arrived, merge the trajectory data with the same radar id and vehicle id, and generate the trajectory id with the data timestamp of the starting time point. The format of the trajectory id is: radar id, vehicle id and starting timestamp. At the same time, set the matching identification field corresponding to the trajectory id. The matching identification field is divided into start_matched and end_matched, and the values ​​are true or false respectively; put the trajectory data, the corresponding trajectory id and the matching identification field into S all Among them, S all Includes all the trajectories that have been tracked and reached at all intersections.

[0078] Two independent trajectory pools are set at the intersection, namely the entrance trajectory pool S in and exit trajectory pool S out When assigning a trajectory, let the threshold σ 1 = 60m, calculate the distance from each intersection to the starting point of the trajectory, if the smallest distance is less than σ 1 Then the trajectory is assigned to the exit trajectory pool S of the intersection closest to the starting point of the trajectory out . Calculate the distance between each intersection and the end point of the trajectory. If the minimum distance is less than σ 1 The trajectory is assigned to the entry trajectory pool of the intersection closest to the end point of the trajectory. Since there is only one intersection in this embodiment, it is only necessary to determine whether the distance between the starting point and the end point of the trajectory and the intersection is less than the threshold σ 1 That's it.

[0079] Traverse each intersection in turn, if the entrance trajectory pool S of the intersection in and exit trajectory pool S out Calculate the entrance trajectory pool S of the intersection in and exit trajectory pool S out The distance between two pairs of trajectories. When calculating, given the entrance trajectory pool S in Trajectory and exit trajectory pool S out The trajectory in in is a trajectory point, including latitude, longitude and timestamp. First, determine the two trajectories T 1 and T 2 Are the radar ids in the trajectory ids the same? If they are the same, let T 1 With T 2 If the trajectory distance between them is positive infinity, the calculation ends; if it is different, T 1 In T2 Extrapolate on timestamps where time is greater than Less than The fragments are filled using linear interpolation, less than The fragment is extended by linear interpolation of the first two trajectory points, which is greater than The segment is extended by linear interpolation of the two ending trajectory points to obtain the extended trajectory T 1 ′; will be greater than The epitaxial fragment of T is rotated 45°, 90°, -45° and -90° to obtain 1 ′ a ,T 1 ′ b ,T 1 ′ c ,T 1 ′ d Four extended trajectories simulating turns; calculate T 1 ′ ,T 1 ′ a ,T 1 ′ b ,T 1 ′ c ,T 1 ′ d Respectively with T 2 The smallest value is taken as the final trajectory distance.

[0080] Arrange the trajectory distances of all paired situations at the same intersection in ascending order, and take out the corresponding trajectory pairs from small to large. Set the threshold σ 2 = 20m, if the track distance is less than σ 2 , and the trajectories in the trajectory pair have not been successfully matched by other trajectories in this matching, then it means that the trajectory pair is successfully matched. Then the successfully matched trajectory pair is selected from the exit trajectory pool S out The trajectory taken out sets start_matched = true matching flag, and sets its global id to the global id of another trajectory it matches, and adds it to the entry trajectory pool S in The trajectory taken out is set to end_matched = true matching flag. Finally, the successfully matched trajectory pairs are cleared out from the two trajectory pools at the intersection.

[0081] Traverse the global trajectory pool S allIf there is a trajectory that has not been matched for a long time, it will be cleared out of the global trajectory pool. For the trajectory that has been successfully matched, the trajectory will be cleared and the trajectory that matches it will be cleared according to its global id.

[0082] It can be seen from the above technical scheme that the vehicle multi-section trajectory matching method and system based on radar detection provided by the embodiment of the present invention sets an entry pool and an exit pool at the intersection, calculates the distance of the trajectory pairs in the entry pool and the exit pool, and performs real-time matching on the same vehicle detected by different radars, so as to realize that the vehicle travels from one section to another, and the ID under different radars remains unchanged, and the complete driving trajectory of the vehicle is obtained. The present invention is simple to implement, has loose data quality requirements, and is flexible to implement. It only needs to calibrate the central longitude and latitude of each intersection, and does not need to calibrate other information such as the lane range, the allowed driving direction of the lane, and the upstream and downstream adjacency relationship of the section; it does not require that the coverage ranges of different radars overlap, and the requirements for data quality are not harsh; when implemented, it can be deployed for a road network containing hundreds of intersections or for several adjacent or non-adjacent intersections. In places where the radar coverage ranges intersect but there are no intersections, intersections can be manually added to complete the deployment. It has high flexibility, a wide range of application, and strong practical ability.

[0083] The above description is only a preferred embodiment of the present invention and an explanation of the technical principles used. It is not intended to limit the scope of the invention claimed for protection, but only represents the preferred embodiment of the present invention. Those skilled in the art should understand that the scope of the invention involved in the present invention is not limited to the technical solution formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the inventive concept. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present invention.

Claims

1. A vehicle multi-section trajectory matching method based on radar detection, characterized in that: The method comprises the following steps: Step S1, obtaining the track data of the intersection detected by the radar, and when the original track data detected by the radar is not read after a predetermined time, it is determined that the track data detected by the radar for the vehicle has all arrived; Step S2, establish a global trajectory pool S all ; Preprocess all the trajectory data that have arrived, obtain the trajectory id, and set the matching identification field corresponding to the trajectory id; and put all the trajectory data, the corresponding trajectory id and the matching identification field into S all Among them, S all The trajectories of all intersections tracked and reached are included; the matching identification field is divided into start_matched and end_matched, and the values ​​are true or false respectively; Step S3: two independent trajectory pools are set at each intersection, namely, the entrance trajectory pool S in and exit trajectory pool S out , S all The trajectories in the S are assigned to the corresponding intersections according to the predetermined logic. in or S out middle; Step S4, traverse each intersection in turn, and compare the S in and S out Pair the trajectories in pairs, calculate the corresponding trajectory distances, and sort all calculated trajectory distances at the same intersection in ascending order, and take out the trajectory pairs whose trajectory distance is less than the threshold σ2; Step S5, identifying the trajectory data in the extracted trajectory pair, and if the trajectory data has not been matched before, determining that the trajectory pair is matched successfully; Step S6, the matched trajectory is aligned, and the out Set start_matched = true to match the trajectory extracted from S in The trajectory taken from the trajectories is set to end_matched = true matching flag; the successful matching result is recorded as output, and the successful matching trajectory pairs are respectively taken from the S of the intersection in and S out Clean it out.

2. The vehicle multi-segment trajectory matching method based on radar detection according to claim 1, characterized in that: The predetermined logic in step S3 includes the following steps: Step S31, if the matching flag of the trajectory is start_matched = true, then skip the current trajectory; otherwise, execute the following steps: take the intersection center as the intersection coordinate value, calculate the distance of each intersection from the starting point of the trajectory data corresponding to the current trajectory id and sort them, if the distance between the starting point and the nearest intersection is less than the threshold σ1, then assign the trajectory to the intersection closest to the starting point S out , otherwise skip; Step S32, if the matching flag of the trajectory end_matched = true, skip the current trajectory; otherwise, execute the following steps: calculate the distance between each intersection and the end point of the trajectory data corresponding to the current trajectory id and sort them. If the distance between the end point and the nearest intersection is less than the threshold σ1, assign the trajectory to the intersection closest to the end point. in , otherwise skip; Step S33: If a trajectory is not assigned to any intersection, in or S out , the vehicle is considered to have left the road network in the middle of a certain road section and no operation is performed on the vehicle.

3. The vehicle multi-segment trajectory matching method based on radar detection according to claim 1, characterized in that: When calculating the corresponding trajectory distance in step S4, both the time and space information of the trajectory are considered.

4. The vehicle multi-segment trajectory matching method based on radar detection according to claim 3 is characterized in that: The specific process of calculating the corresponding trajectory distance is as follows: Step S41, given two trajectories in is the i-th trajectory point of the j-th trajectory, are the longitude, latitude and timestamp of the i-th track point of the j-th track respectively; Step S42, determining whether the radar IDs in the track IDs of the two tracks T1 and T2 are the same, if they are the same, setting the track distance between T1 and T2 to positive infinity and ending the calculation; if they are different, executing step S43; Step S43, extrapolate T1 to the timestamp of T2, where the time is greater than Less than The fragments are filled using linear interpolation, less than The fragment is extended by linear interpolation of the first two trajectory points, which is greater than The segment is extended by linear interpolation of the two ending trajectory points to obtain the extended trajectory T1′; Step S44, greater than The epitaxial fragment is rotated 45°, 90°, -45° and -90° to obtain T′ 1a ,T′ 1b ,T′ 1c ,T′ 1d Four extended trajectories simulating turns; Step S45, calculate T'1, T' 1a ,T′ 1b ,T′ 1c ,T′ 1d The smallest value is taken as the trajectory distance between T1 and T2.

5. The vehicle multi-segment trajectory matching method based on radar detection according to claim 1, characterized in that: When acquiring radar detection data, every τ input s reads the trajectory input generated during the monitored time period, τ input The time interval for reading data.

6. The vehicle multi-segment trajectory matching method based on radar detection according to claim 1, characterized in that: The fields contained in the original trajectory data in step S1 include: radar id, vehicle id, timestamp, longitude, latitude and speed.

7. The vehicle multi-segment trajectory matching method based on radar detection according to claim 1, characterized in that: The trajectory ID includes a radar ID, a vehicle ID and a start timestamp.

8. The vehicle multi-segment trajectory matching method based on radar detection according to claim 1, characterized in that: Step S6 also includes: at a preset time point, traversing S all If there is a trajectory that is not matched within the predetermined time, the trajectory is removed from the global trajectory pool S all Clean it out.

9. A vehicle multi-section trajectory matching system based on radar detection, characterized in that: The system includes: a data acquisition module, a global track pool establishment module, an independent track pool setting module, a track allocation module, a track distance calculation module, a matching recognition module, and a post-processing module; wherein, The data acquisition module is used to acquire the trajectory data of the intersection detected by the radar. When the original trajectory data detected by the radar is not read after a predetermined period of time, it is determined that the trajectory data detected by the radar for the vehicle has all arrived; The global trajectory pool establishment module is used to establish a global trajectory pool S all ; Preprocess all the trajectory data that have arrived, obtain the trajectory id, and set the matching identification field corresponding to the trajectory id; and put all the trajectory data, the corresponding trajectory id and the matching identification field into S all Among them, S all The trajectories of all intersections tracked and reached are included; the matching identification field is divided into start_matched and end_matched, and the values ​​are true or false respectively; The independent track pool setting module is used to set two independent track pools at each intersection, namely the entrance track pool S in and exit trajectory pool S out ; The trajectory allocation module is used to allocate S all The trajectories in the S are assigned to the corresponding intersections according to the predetermined logic. in or S out middle; The trajectory distance calculation module is used to traverse each intersection in turn and calculate the S in and S out Pair the trajectories in pairs, calculate the corresponding trajectory distances, and sort all calculated trajectory distances at the same intersection in ascending order, and take out the trajectory pairs whose trajectory distance is less than the threshold σ2; The matching identification module is used to identify the trajectory data in the extracted trajectory pair, and if the trajectory data has not been matched before, it is determined that the trajectory pair is matched successfully; The post-processing module is used to center the successfully matched trajectory from S out Set start_matched = true to match the trajectory extracted from S in The trajectory taken from the trajectories is set to end_matched = true matching flag, the successful matching result is recorded as output, and the successful matching trajectory pairs are respectively taken from the intersection S in and S out Clean it out.

10. The vehicle multi-segment trajectory matching system based on radar detection according to claim 9, characterized in that: The trajectory distance calculation module considers both the time and space information of the trajectory when calculating the corresponding trajectory distance.

Images