A vehicle GPS trajectory prediction method, system, device and medium based on road network location information

Abstract

The present invention relates to a vehicle GPS trajectory prediction method, system, device and medium based on road network position information, and the method includes: extracting model data from map data and processing to form a road network model; receiving and tracking real-time GPS coordinate data of the vehicle to determine whether the vehicle trajectory is interrupted; if the vehicle trajectory is interrupted, for the blind spot vehicle whose position is not updated and no latest GPS coordinate data is transmitted, converting the historical GPS coordinates of the blind spot vehicle into the projection position in the road network model; calculating the speed of the blind spot vehicle in the previous period according to the following model and the projection position, searching for the front vehicle and calculating the speed difference and position difference between the front and rear vehicles, updating the acceleration and speed of the blind spot vehicle to determine the road network predicted position of the blind spot vehicle; mapping the road network predicted position of the blind spot vehicle to the GPS coordinate form, updating the blind spot vehicle coordinate point, and outputting the latest position of the vehicle. The present invention provides a more accurate, comprehensive and adaptable vehicle trajectory prediction solution.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle flow trajectory prediction, and in particular to a vehicle GPS trajectory prediction method, system, device and medium based on road network location information. Background Art

[0002] In today's highly intelligent transportation field, it is particularly important to accurately obtain and predict the location and trajectory of vehicles. Road test radar and monitoring systems have been widely used as key tools to collect vehicle GPS data, and they play an indispensable role in the road network. However, in practical applications, these systems face many challenges.

[0003] Due to the presence of objects blocking the real environment, the transmission and reception of radar signals may be disturbed, thus affecting the accuracy of the data. In addition, if the intervals between radar and monitoring equipment are too large, some areas may not be effectively covered, resulting in the loss of vehicle location information. At the same time, network fluctuations are also an issue that cannot be ignored. It may cause delays, losses, or even interruptions in data transmission, seriously affecting the real-time and accuracy of vehicle positioning. These problems together constitute major challenges in vehicle accurate positioning, efficient navigation, and scientific traffic management.

[0004] There are currently many methods for predicting vehicle trajectories, but each method has its limitations. Although the linear or polynomial fitting method based on historical data can calculate the speed, distance and other information of the vehicle in combination with historical data, and predict the position through kinematic formulas, this method may not accurately capture the nonlinear characteristics of vehicle motion in the case of traffic congestion or frequent speed changes of vehicles, resulting in a decrease in prediction accuracy. At the same time, this method does not take into account the geographical information of the road network, so the prediction results under special road conditions such as curves may not be in line with reality.

[0005] The vehicle trajectory prediction method based on artificial intelligence relies on a large amount of historical vehicle data to train the prediction model. However, the comprehensiveness and quality of the data have a significant impact on the model effect. In addition, the high computing resource consumption of technologies such as deep learning makes the deployment and operation costs high. At the same time, the generalization ability of the model, adaptability to new scenarios, and convenience of updating and maintenance are also challenges faced by this method.

[0006] The real-time online vehicle traffic simulation method builds a traffic simulation road network model to deduce the vehicle motion state. However, this method focuses more on the simulation of trajectories rather than the display of real trajectory points, and has a high time delay because it requires the vehicle to pass through a specific node before trajectories between nodes can be backtracked and simulated. These limitations restrict the application of this method in real-time precise positioning. Summary of the invention

[0007] 1. Technical issues to be resolved

[0008] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a vehicle GPS trajectory prediction method, system, device and medium based on road network location information, which solves the technical problems faced by the prior art in acquiring and predicting vehicle position and trajectory, such as object occlusion, insufficient equipment coverage, network fluctuations, limitations of prediction methods, strong data dependence, high computing cost, weak generalization ability and time delay.

[0009] (II) Technical solution

[0010] In order to achieve the above object, the main technical solutions adopted by the present invention include:

[0011] In a first aspect, an embodiment of the present invention provides a vehicle GPS trajectory prediction method based on road network location information, comprising:

[0012] Extracting model data from the acquired map data and processing it to form a road network model;

[0013] Receive and track the real-time GPS coordinate data of the detected vehicle, and determine whether the vehicle trajectory is interrupted based on the timestamp;

[0014] If the vehicle trajectory is interrupted, for the blind spot vehicles whose positions are not updated and the latest GPS coordinate data is not transmitted, the historical GPS coordinates of the blind spot vehicles are converted into the projected positions in the road network model through step-by-step retrieval and projection calculation;

[0015] The speed of the vehicle in the blind spot in the previous period is calculated based on the following model and the projected position, the front vehicle is searched and the speed difference and position difference between the front and rear vehicles are calculated, and the acceleration and speed of the vehicle in the blind spot are updated to determine the predicted road network position of the vehicle in the blind spot;

[0016] The predicted road network position of the blind spot vehicle is mapped to GPS coordinates, and the blind spot vehicle coordinate points are updated to output the latest vehicle position.

[0017] Optionally, extracting model data from the acquired map data and processing to form a road network model includes:

[0018] Identify and extract lane centerline data including a lane centerline number, a lane number, and a set of longitude and latitude coordinate points from a starting point to an end point from the acquired map data;

[0019] Identify and extract lane geographic shape information including lane number, road section number, and longitude and latitude point set from the acquired map data;

[0020] Identify and extract the road segment geographic shape information including the road segment number, the road number to which it belongs, and the set of longitude and latitude coordinate points from the acquired map data;

[0021] Identify and extract road geographic shape information including road numbers and longitude and latitude coordinate point sets from the acquired map data;

[0022] Standardize lane centerline data, lane surface data, section surface data, and road surface data, and record the upstream and downstream relationships and subordinate relationships between each data;

[0023] A road network model is formed based on the standardized data and the upstream, downstream and subordinate relationships between the various data.

[0024] Optionally, receiving and tracking the real-time GPS coordinate data of the detected vehicle, and determining whether the vehicle track is interrupted according to the timestamp includes:

[0025] Receive real-time GPS coordinate data of the vehicle detected by the detection equipment;

[0026] Using the target ID as the unique identifier of each vehicle, the vehicle's location history data is tracked at fixed time intervals and integrated into a set of historical location points for each vehicle, where each location point contains the vehicle's latitude and longitude information and the timestamp of the corresponding point;

[0027] Extract the latest location point data from the historical location point set of each vehicle;

[0028] Compare the timestamp of the latest location point data with the current time and calculate the time difference;

[0029] If the calculated time difference is greater than the preset threshold, the current vehicle trajectory is determined to be interrupted, indicating that the vehicle position information at the current time is missing, and the next processing flow is entered;

[0030] If the calculated time difference is less than or equal to the preset threshold, it is determined that the vehicle detection trajectory is not interrupted and the latest position of the vehicle is directly output.

[0031] Optionally, if the vehicle trajectory is interrupted, for the blind spot vehicle whose position is not updated and no latest GPS coordinate data is transmitted, the historical GPS coordinates of the blind spot vehicle are converted into the projected position in the road network model through step-by-step retrieval and projection calculation, including:

[0032] Identify and select vehicles whose positions have not been updated and whose latest GPS coordinate data has not been transmitted as blind spot vehicles that need to be processed;

[0033] Using the historical GPS coordinates of the vehicle in the blind spot and combining it with the geographical shape information of the road, the road to which the historical GPS coordinates of the vehicle in the blind spot belong is retrieved in the road network model, and the corresponding road identifier is recorded;

[0034] According to the road identifier, all the road sections included in the corresponding road are screened out from the preset road section information table, and the road section where the blind spot vehicle coordinate point is located is retrieved in the road network model in combination with the road section geographic shape information, and the corresponding road section identifier is recorded;

[0035] According to the road segment identifier, all lanes included in the corresponding road segment are screened from the preset lane information table, and the lane where the coordinate point of the blind spot vehicle is located is retrieved from the road network model in combination with the lane geographic shape information, and the corresponding lane identifier is recorded;

[0036] According to the lane identifier, extracting the corresponding lane centerline represented in the form of a series of broken line segments from a preset lane centerline information table;

[0037] For each coordinate point of a vehicle in the blind spot, the following operations are repeated until all line segments of the lane centerline are processed: a line segment is selected from a preset polyline shape list of the lane centerline, and based on the projection relationship between the coordinate point of the vehicle in the blind spot and the selected line segment, whether the projection point is inside the line segment is determined, and if it is inside the line segment, the line segment is marked;

[0038] Among all the marked line segments, select a line segment that minimizes the distance between the blind spot vehicle coordinate point and the line segment as the projection target line segment;

[0039] According to the index of the projected target line segment, the accumulated distance from the starting point of the line segment to the starting point of the lane centerline is obtained, so as to determine the projection position of the blind spot vehicle in the road network model.

[0040] Optionally, a line segment is selected from a preset polyline shape list of the lane centerline, and based on the projection relationship between the blind spot vehicle coordinate point and the selected line segment, it is determined whether the projection point is located inside the line segment. If it is located inside the line segment, marking the line segment includes:

[0041] Select a line segment AB from the preset polyline shape list of the lane centerline as the current processing object;

[0042] Determine the coordinates of the starting point A and the end point B of the current processing object and the coordinate point P of the blind spot vehicle, and obtain the first vector from the starting point A of the line segment to the coordinate point P of the blind spot vehicle , and a second vector from the starting point A to the end point B of the line segment ;

[0043] According to the first vector and the second vector The angle between them is used to find the first vector In the second vector The projection length on the second vector is divided by the projection length The scalar length of the coordinate point P is obtained by The projection relationship;

[0044] Based on the projection relationship, determine whether the projection point of the coordinate point P of the blind spot vehicle is within the selected line segment AB;

[0045] If the value of the projection relationship is within the set interval, the projection point of the coordinate point P is determined to be within the line segment AB, and the line segment is marked;

[0046] If the value of the projection relationship is not within the set interval, it is determined that the projection point of the coordinate point P is not within the line segment AB, and the line segment is skipped and the next line segment is found as the current processing object for processing.

[0047] Optionally, calculating the speed of the blind spot vehicle in the previous period according to the following model and the projection position, searching for the front vehicle and calculating the speed difference and position difference between the front and rear vehicles, and updating the acceleration and speed of the blind spot vehicle to determine the road network predicted position of the blind spot vehicle includes:

[0048] Calculate the speed of the vehicle in the blind spot during the last set period of time according to the distance traveled by the vehicle in the blind spot during the last set period of time;

[0049] According to the projection position of the blind spot vehicle in the road network model, in the same lane, all other vehicles whose cumulative distance from the starting point is greater than the projection position of the blind spot vehicle in the road network model are screened out, and the vehicle closest to the blind spot vehicle is selected and marked as the front vehicle;

[0050] Calculate the speed difference and position difference between the blind spot vehicle and the vehicle in front;

[0051] Apply the intelligent driving formula in the car-following model, combine the speed difference, position difference, vehicle expected speed, maximum acceleration, comfortable deceleration, acceleration coefficient, minimum safety distance and safe headway time to calculate the expected safe distance and acceleration of vehicles in the blind spot;

[0052] The speed of the current period is obtained based on the calculated acceleration, and the speed of the vehicle in the blind spot during the previous set period is updated to obtain the distance traveled during the current period.

[0053] Determine the predicted road network position of the blind spot vehicle in the current period based on the projection position of the blind spot vehicle on the lane in the previous period and the distance it has traveled in the current period;

[0054] in,

[0055] The intelligent driving formula is:

[0056] ;

[0057] In the formula, is the expected safe distance between vehicles, v n The speed of the vehicle in the blind spot during the last set period of time The speed difference between the vehicle in the blind spot and the vehicle in front, S 0 is the minimum safety distance, T For safe headway, a is its maximum acceleration, b For comfortable deceleration, a n is the acceleration, v 0 is the desired speed of the vehicle, is the acceleration coefficient, It is the position difference between the blind spot vehicle and the front vehicle.

[0058] Optionally, the predicted road network position of the blind spot vehicle is mapped to GPS coordinates, and the blind spot vehicle coordinates are updated. The output of the latest vehicle position includes:

[0059] Compare the predicted road network position of the blind spot vehicle with the accumulated distance from each node of the lane centerline to the starting point to find the index of the line segment where the blind spot vehicle is located i ;

[0060] According to the determined line segment index i , get the line segment from the list of preset polyline shapes for lane centerlines i The starting and ending coordinates of the vehicle are calculated by the conversion formula to obtain the GPS coordinates of the vehicle in the blind spot.

[0061] The calculated GPS coordinates of the blind spot vehicle are used as the vehicle's latest current position, and the coordinate point information of the blind spot vehicle is updated and output for subsequent use or display;

[0062] in,

[0063] The conversion formula is:

[0064] ;

[0065] In the formula, Predict the road network position of vehicles in the blind spot, is the coordinate of the starting point of the line segment where the vehicle in the blind spot is located, is the coordinate of the end point of the line segment where the vehicle in the blind spot is located, The predicted road network position of the vehicle in the blind spot during the current period. , They are respectively the distances from the starting point and end point of the line segment where the vehicle in the blind spot is located to the starting point of the lane centerline of the line segment.

[0066] In a second aspect, an embodiment of the present invention provides a vehicle GPS trajectory prediction system based on road network location information, comprising:

[0067] A model forming module, used for extracting model data from the acquired map data and processing it to form a road network model;

[0068] The track judgment module is used to receive and track the real-time GPS coordinate data of the detected vehicle and judge whether the vehicle track is interrupted according to the timestamp;

[0069] The position projection module is used to convert the historical GPS coordinates of the blind spot vehicle into the projection position in the road network model through step-by-step retrieval and projection calculation if the vehicle trajectory is interrupted and the position of the blind spot vehicle is not updated and the latest GPS coordinate data is not transmitted;

[0070] The position prediction module is used to calculate the speed of the blind spot vehicle in the previous period based on the following model and the projected position, search for the leading vehicle and calculate the speed difference and position difference between the leading and trailing vehicles, and update the acceleration and speed of the blind spot vehicle to determine the road network predicted position of the blind spot vehicle;

[0071] The position update module is used to map the road network predicted position of the blind spot vehicle into GPS coordinate form, update the blind spot vehicle coordinate point, and output the latest vehicle position.

[0072] In a third aspect, an embodiment of the present invention provides a vehicle GPS trajectory prediction device based on road network location information, comprising:

[0073] at least one database;

[0074] and a memory in communication with at least one database;

[0075] The memory stores instructions that can be executed by at least one database, and the instructions are executed by at least one database so that the at least one database can execute the vehicle GPS trajectory prediction method based on road network location information as described above.

[0076] In a fourth aspect, an embodiment of the present invention provides a computer-readable medium having computer executable instructions stored thereon, and when the executable instructions are executed by a processor, the vehicle GPS trajectory prediction method based on road network location information as described above is implemented.

[0077] (III) Beneficial effects

[0078] The beneficial effects of the present invention are as follows: the present invention not only overcomes many defects of the prior art, but also provides a more accurate, comprehensive and adaptable trajectory prediction solution. First, the key model data is extracted from the acquired map data and processed carefully to construct a detailed road network model, providing a solid foundation for subsequent trajectory prediction. Secondly, the real-time GPS coordinate data of the detected vehicle is received and continuously tracked, and the vehicle trajectory is accurately judged based on the timestamp whether there is an interruption.

[0079] Once the vehicle trajectory is interrupted, that is, for those blind spot vehicles whose positions are not updated in time and no latest GPS coordinate data is transmitted, the present invention converts the historical GPS coordinates of the blind spot vehicle into a specific projection position in the road network model through efficient step-by-step retrieval and accurate projection calculation. Next, the speed of the blind spot vehicle in the previous period is accurately calculated using the advanced car-following model and projection position data. Furthermore, by searching for the leading vehicle and calculating the speed difference and position difference between the leading and trailing vehicles, the acceleration and speed of the blind spot vehicle can be updated in real time, thereby accurately determining the predicted position of the blind spot vehicle in the road network.

[0080] Finally, the present invention accurately maps the predicted position of the blind spot vehicle in the road network back to the GPS coordinate form, timely updates the coordinate points of the blind spot vehicle, and accurately outputs the latest position of the vehicle. This series of processes not only ensures the accuracy and reliability of trajectory prediction, but also enables the present invention to perform well when dealing with short-term data loss, ensuring the continuity of the trajectory and the stability of the system. By integrating multi-dimensional information and adopting a prediction model that adapts to nonlinear characteristics, the present invention provides more accurate, timely and effective support for traffic management and navigation. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 A schematic flow chart of a vehicle GPS trajectory prediction method based on road network location information is provided for an embodiment of the present invention;

[0082] Figure 2 A specific flow chart of step S1 of a method for predicting vehicle GPS trajectory based on road network location information is provided for an embodiment of the present invention;

[0083] Figure 3 A specific flow chart of step S2 of a method for predicting vehicle GPS trajectory based on road network location information is provided for an embodiment of the present invention;

[0084] Figure 4 A specific flow chart of step S3 of a method for predicting vehicle GPS trajectory based on road network location information is provided for an embodiment of the present invention;

[0085] Figure 5 A specific flow chart of step S36 of a method for predicting vehicle GPS trajectory based on road network location information is provided for an embodiment of the present invention;

[0086] Figure 6 A schematic diagram of trajectory point projection of a vehicle GPS trajectory prediction method based on road network location information is provided for an embodiment of the present invention;

[0087] Figure 7 A specific flow chart of step S4 of a method for predicting vehicle GPS trajectory based on road network location information is provided for an embodiment of the present invention;

[0088] Figure 8 A simulation schematic diagram of a vehicle GPS trajectory prediction method based on road network location information is provided for an embodiment of the present invention;

[0089] Fig. 9 A specific flow chart of step S5 of a method for predicting vehicle GPS trajectory based on road network location information is provided for an embodiment of the present invention;

[0090] Fig.10 The present invention provides an overall flow chart of a vehicle GPS trajectory prediction method based on road network location information. DETAILED DESCRIPTION

[0091] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0092] Prior to this, in order to facilitate understanding of the technical solution provided by the present invention, some concepts are first introduced below.

[0093] Following model: In traffic flow, the following vehicle adjusts its driving behavior according to the status of the leading vehicle (such as speed, acceleration, position, etc.), including acceleration, deceleration or maintaining a constant speed.

[0094] Safe headway: The time interval between two vehicles passing the same location in traffic flow.

[0095] like Figure 1 As shown, a vehicle GPS trajectory prediction method based on road network position information proposed in an embodiment of the present invention includes: extracting model data from the acquired map data and processing it to form a road network model; receiving and tracking the real-time GPS coordinate data of the detected vehicle, and judging whether the vehicle trajectory is interrupted according to the timestamp; if the vehicle trajectory is interrupted, for the blind spot vehicle whose position has not been updated and no latest GPS coordinate data has been transmitted, the historical GPS coordinates of the blind spot vehicle are converted into the projection position in the road network model through step-by-step retrieval and projection calculation; the speed of the blind spot vehicle in the previous period is calculated according to the following model and the projection position, the front vehicle is searched and the speed difference and position difference between the front and rear vehicles are calculated, and the acceleration and speed of the blind spot vehicle are updated to determine the road network predicted position of the blind spot vehicle; the road network predicted position of the blind spot vehicle is mapped to the GPS coordinate form, and the blind spot vehicle coordinate point is updated to output the latest position of the vehicle.

[0096] The present invention not only overcomes many defects of the prior art, but also provides a more accurate, comprehensive and adaptable trajectory prediction solution. First, the key model data is extracted from the acquired map data and processed carefully to construct a detailed road network model, providing a solid foundation for subsequent trajectory prediction. Secondly, the real-time GPS coordinate data of the detected vehicle is received and continuously tracked, and the vehicle trajectory is accurately judged based on the timestamp whether it is interrupted.

[0097] Once the vehicle trajectory is interrupted, that is, for those blind spot vehicles whose positions are not updated in time and no latest GPS coordinate data is transmitted, the present invention converts the historical GPS coordinates of the blind spot vehicle into a specific projection position in the road network model through efficient step-by-step retrieval and accurate projection calculation. Next, the speed of the blind spot vehicle in the previous period is accurately calculated using the advanced car-following model and projection position data. Furthermore, by searching for the leading vehicle and calculating the speed difference and position difference between the leading and trailing vehicles, the acceleration and speed of the blind spot vehicle can be updated in real time, thereby accurately determining the predicted position of the blind spot vehicle in the road network.

[0098] Finally, the present invention accurately maps the predicted position of the blind spot vehicle in the road network back to the GPS coordinate form, timely updates the coordinate points of the blind spot vehicle, and accurately outputs the latest position of the vehicle. This series of processes not only ensures the accuracy and reliability of trajectory prediction, but also enables the present invention to perform well when dealing with short-term data loss, ensuring the continuity of the trajectory and the stability of the system. By integrating multi-dimensional information and adopting a prediction model that adapts to nonlinear characteristics, the present invention provides more accurate, timely and effective support for traffic management and navigation.

[0099] In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0100] Specifically, a vehicle GPS trajectory prediction method based on road network location information includes:

[0101] S1. Extracting model data from the acquired map data and processing it to form a road network model.

[0102] The embodiment of the present invention uses vector high-precision map data provided by professional map manufacturers. This data has the characteristics of high precision and high information content, and can reflect the real situation of the road network in detail, including rich information such as road direction, width, traffic signs, and traffic lights.

[0103] It should be noted that the present invention also provides a corresponding solution for scenarios without high-precision maps. In this case, trajectory prediction will degenerate into a solution that only performs linear prediction based on the vehicle's historical trajectory. Although the accuracy of this solution may not be comparable to that of a solution based on high-precision maps, it can provide certain prediction capabilities in scenarios where high-precision map data is lacking.

[0104] Furthermore, if Figure 2 As shown, step S1 includes:

[0105] S11. Identify and extract lane centerline data including a lane centerline number, a lane number, and a set of longitude and latitude coordinate points from a starting point to an end point from the acquired map data.

[0106] S12. Identify and extract lane geographic shape information including lane number, road section number, and longitude and latitude point set from the acquired map data.

[0107] S13, identifying and extracting the road segment geographic shape information including the road segment number, the road number to which it belongs, and the set of longitude and latitude coordinate points from the acquired map data.

[0108] S14. Identify and extract road geographic shape information including road numbers and longitude and latitude coordinate point sets from the acquired map data.

[0109] S15. Standardize the lane centerline data, lane surface data, road section surface data, and road surface data, and record the upstream and downstream relationships and subordinate relationships between the various data.

[0110] S16. A road network model is formed based on the standardized data and the upstream and downstream relationships and subordinate relationships between the various data.

[0111] In one embodiment, the present invention extracts a digital twin model of the actual road network from high-precision map data, including lane centerline Line, lane surface data Lane, section surface data Edge, and road surface data Road. Its input format needs to be standardized, and the upstream and downstream relationships within each level of data such as lane centerline, lane surface, section surface, and road surface, as well as the subordinate relationships between different levels of data, need to be recorded.

[0112] RoadRoad From the starting point to the end point, it is composed of multiple road sections connected to each other, covering the entire road surface. It includes the road number and the road surface range composed of a set of longitude and latitude coordinate points.

[0113] The road segment edge usually contains a continuous small section of road surface with no feature changes in the road channelization in the road model, which is composed of multiple lane surfaces in the same direction. It contains the road segment number, the road number to which it belongs, and the road segment surface range composed of a set of longitude and latitude coordinate points.

[0114] Lanes belong to road sections and are usually split into upstream and downstream lanes at the point where the road channelization changes. The lane surface range consists of the lane number, the road section number to which it belongs, and the longitude and latitude point set.

[0115] The lane centerline Line corresponds to the lane one by one, and is formed by connecting multiple polylines in sequence, including the lane centerline number, the lane number to which it belongs, and the longitude and latitude coordinate point set from the starting point to the end point of the lane line. When extracting, the cumulative distance between the node of each polyline segment and the starting point of the lane is calculated to facilitate the subsequent determination of its position in the road network when projecting the vehicle's GPS coordinate point.

[0116] S2. Receive and track the real-time GPS coordinate data of the detected vehicle, and determine whether the vehicle track is interrupted based on the timestamp.

[0117] Furthermore, if Figure 3 As shown, step S2 includes:

[0118] S21, receiving real-time GPS coordinate data of the vehicle detected by the detection equipment.

[0119] S22. Using the target ID as the unique identifier of each vehicle, track the vehicle's location history data at fixed time intervals and integrate it into a set of historical location points for each vehicle, where each location point contains the latitude and longitude information of the vehicle and the timestamp of the corresponding point.

[0120] S23. Extract the latest location point data from the historical location point set of each vehicle.

[0121] S24. Compare the timestamp of the latest location point data with the current time and calculate the time difference.

[0122] S25. If the calculated time difference is greater than a preset threshold, it is determined that the current vehicle trajectory is interrupted, indicating that the vehicle position information at the current time is missing, and the next processing flow is entered.

[0123] S26. If the calculated time difference is less than or equal to a preset threshold, it is determined that the vehicle detection track is not interrupted, and the latest position of the vehicle is directly output.

[0124] In one embodiment, the present invention receives real-time coordinate data of vehicles detected by roadside radar, surveillance cameras and other equipment, and uniquely identifies the target ID and sends it to the target ID at a fixed time interval. , tracking the historical location data of the vehicle. The historical location point set of each vehicle N is , each location point has the latitude and longitude information of the vehicle In addition, it also contains the detection timestamp of the corresponding point t i. Get the latest location of the vehicle p n ,like p n The corresponding timestamp t n The time difference with the current time is greater than , it means that the current vehicle trajectory is interrupted and the vehicle position information at the current time is missing, so go to the next step. Otherwise, the vehicle detection trajectory is normal, skip the processing, and directly output the current latest position p n .

[0125] S3. If the vehicle trajectory is interrupted, for the blind spot vehicle whose position is not updated and no latest GPS coordinate data is transmitted, the historical GPS coordinates of the blind spot vehicle are converted into the projection position in the road network model through step-by-step retrieval and projection calculation.

[0126] The present invention selects vehicles whose positions have not been updated and whose latest GPS coordinates have not been transmitted, and converts the historical GPS coordinates of the vehicles into Converted into a location form in the road network, including the road, road section, lane to which the location point belongs, and the cumulative distance from the starting point of the lane.

[0127] Furthermore, if Figure 4 As shown, step S3 includes:

[0128] S31. Identify and select vehicles whose positions have not been updated and whose latest GPS coordinate data has not been transmitted as blind spot vehicles that need to be processed.

[0129] S32. Using the historical GPS coordinates of the vehicle in the blind spot and combining them with the geographical shape information of the road, the road to which the historical GPS coordinates of the vehicle in the blind spot belong is retrieved in the road network model, and the corresponding road identifier is recorded.

[0130] S33. According to the road identifier, all road sections included in the corresponding road are screened out from the preset road section information table, and the road section where the blind spot vehicle coordinate point is located is retrieved from the road network model in combination with the road section geographic shape information, and the corresponding road section identifier is recorded.

[0131] S34. According to the road section identifier, all lanes included in the corresponding road section are screened out from the preset lane information table, and the lane where the blind spot vehicle coordinate point is located is retrieved from the road network model in combination with the lane geographic shape information, and the corresponding lane identifier is recorded.

[0132] S35. According to the lane identifier, extract the corresponding lane centerline represented in the form of a series of broken line segments from a preset lane centerline information table.

[0133] In one embodiment, the historical coordinates of the vehicle are known. , retrieve the road where the vehicle coordinate point is located according to the road geographic shape information (surface range) and record the road ID.

[0134] Secondly, according to the road ID, all the road sections included in the road are filtered from the road section information table, and the road section where the vehicle coordinate point is located is retrieved according to the road section geographic shape information (surface range), and the road section ID is recorded.

[0135] Next, according to the road section ID, all lanes contained in the road section are filtered from the lane information table, the lane where the vehicle coordinate point is located is retrieved according to the lane geographic shape information (surface range), and the lane ID is recorded.

[0136] Furthermore, according to the lane ID, the center line of the lane is extracted from the lane center line information table. Since the lane center line is a broken line, the cumulative length pos from the starting point when the vehicle coordinate point is projected on the lane center line is calculated.

[0137] S36. For each coordinate point of the blind spot vehicle, repeat the following operations until all line segments of the lane centerline are processed: select a line segment from the preset broken line shape list of the lane centerline, and determine whether the projection point is inside the line segment based on the projection relationship between the blind spot vehicle coordinate point and the selected line segment; if it is inside the line segment, mark the line segment.

[0138] Furthermore, if Figure 5 As shown, step S36 includes:

[0139] S361. Select a line segment from the preset broken line shape list of the lane centerline as the current processing object.

[0140] S362, determine the coordinates of the starting point A and the end point B of the current processing object and the coordinate point P of the blind spot vehicle, and obtain the first vector from the starting point A of the line segment to the coordinate point P of the blind spot vehicle , and a second vector from the starting point A to the end point B of the line segment .

[0141] S363, according to the first vector and the second vector The angle between them is used to find the first vector In the second vector The projection length on the second vector is divided by the projection length The scalar length of the coordinate point P is obtained by The projection relationship.

[0142] S364. Determine whether the projection point of the coordinate point P of the blind spot vehicle is located within the line segment AB based on the projection relationship.

[0143] S365. If the value of the projection relationship is within the set interval, it is determined that the projection point of the coordinate point P is within the line segment AB, and the line segment is marked.

[0144] S366. If the value of the projection relationship is not within the set interval, it is determined that the projection point of the coordinate point P is not within the line segment AB, and the line segment is skipped and the next line segment is searched for as the current processing object for processing.

[0145] S37. Select a line segment that minimizes the distance between the coordinate point of the vehicle in the blind spot and the line segment from all marked line segments as the projection target line segment;

[0146] S38. According to the index of the projected target line segment, the accumulated distance from the starting point of the line segment to the starting point of the lane centerline is obtained, so as to determine the projected position of the blind spot vehicle in the road network model.

[0147] In another embodiment, reference Figure 6 , the following processing flow is implemented for all line segments of the lane centerline:

[0148] 1) Select the first i Segment line segment, starting point is A , marked as , the end point is marked as B, and the coordinates are , the vehicle position is recorded as P , the coordinates are .

[0149] 2) According to the first vector With the second vector , calculate the angle :

[0150] ;

[0151] 3) Calculate vector In vector Upper projection length:

[0152] ;

[0153] 4) With the projection length , divided by the vector The scalar length of point P is the projection relationship between point P and line segment AB. r :

[0154] ;

[0155] like , then the projection point of point P is within line segment AB, otherwise it means that point P cannot be projected within line segment AB, skip it and find the next segment of the polyline for calculation, mark all line segment.

[0156] 5) Calculate the distance from P to line segment AB and record the corresponding index i :

[0157] ;

[0158] 6) After the vehicle coordinates are projected on all the centerline segments, select all the markers. In the line segment The smallest line segment is used as the projection target line segment, and the index of the line segment is i , read the distance from the starting point of the line segment to the starting point of the lane centerline ,get The projected position of this coordinate point in the road network model. That is, the cumulative length from the starting point POS , so here, POS It is another way to express the projected position of a coordinate point on the center line of the road, and this position is determined by calculating the cumulative length from the starting point of the center line of the road to the projected point.

[0159] S4. Calculate the speed of the blind spot vehicle in the previous period according to the following model and the projection position, search for the front vehicle and calculate the speed difference and position difference between the front and rear vehicles, update the acceleration and speed of the blind spot vehicle, and determine the road network predicted position of the blind spot vehicle.

[0160] Furthermore, if Figure 7 As shown, step S4 includes:

[0161] S41. Calculate the speed of the blind spot vehicle in the last set period of time according to the acquired distance traveled by the blind spot vehicle in the last set period of time.

[0162] S42. According to the projection position of the blind spot vehicle in the road network model, in the same lane, screen out all other vehicles whose cumulative distance from the starting point is greater than the projection position of the blind spot vehicle in the road network model, and select the vehicle closest to the blind spot vehicle from them and mark it as the leading vehicle.

[0163] S43: Calculate the speed difference and position difference between the blind spot vehicle and the preceding vehicle.

[0164] S44. Apply the intelligent driving formula in the car-following model, combine the speed difference, position difference, expected vehicle speed, maximum acceleration, comfortable deceleration, acceleration coefficient, minimum safe distance and safe headway, and calculate the expected safe distance and acceleration of the vehicle in the blind spot.

[0165] S45. Calculate the speed of the current period according to the calculated acceleration, and update the speed of the blind spot vehicle in the previous set period to obtain the distance traveled in the current period.

[0166] S46. Determine the predicted road network position of the blind spot vehicle in the current time period according to the projection position of the blind spot vehicle on the lane in the previous time period and the distance traveled in the current time period.

[0167] In a specific embodiment, the following process is used to calculate and predict the road network position of the vehicle in the blind spot:

[0168] 1) Calculate the speed of the target vehicle in the previous period v n :

[0169] Using the above formula Get the previous time period respectively Internal starting point and end point , and then get the target vehicle N In the previous period Distance travelled within , and then calculate the vehicle's speed in the previous period v n :

[0170] ;

[0171] 2) Search for the previous vehicle:

[0172] According to the projection position of the blind spot vehicle in the road network model , find the lane that satisfies All vehicles, reference Figure 8 , select the vehicle closest to the target The car is the front car.

[0173] 3) Calculate the speed difference between the front and rear vehicles Difference in position (Actual following distance):

[0174] ;

[0175] 4) Update the target vehicle acceleration a n :

[0176] Calculate the expected safe distance between vehicles based on the following model formula (intelligent driving formula) With acceleration a n :

[0177] ;

[0178] In the formula, is the expected safe distance between vehicles, v nThe speed of the vehicle in the blind spot during the last set period of time The speed difference between the vehicle in the blind spot and the vehicle in front, S 0 is the minimum safety distance, the default is 2.5m; T The default time interval for safe headway is 1.5s. a The maximum acceleration of the vehicle, the default value is 2.6m / s 2 ; b For comfortable deceleration, the default value is 4.5m / s 2 ; a n is the acceleration, v 0 The expected speed of the vehicle can be the road speed limit, which is 16.67m / s by default, i.e. 60km / h; is the acceleration coefficient, the default value is 4; It is the position difference between the blind spot vehicle and the front vehicle.

[0179] 5) Update vehicle speed , get the location in the road network :

[0180] Based on the vehicle's previous position on the lane The distance between the vehicle and the wheel , get the position of the vehicle on the lane :

[0181] .

[0182] It should be understood that for the prediction and planning of vehicles in blind spots, only the projection position of the vehicle in the previous round (or at a previous moment) can be obtained. Therefore, the projection position of the blind spot vehicle in the road network model can also be expressed as the position of the blind spot vehicle projected onto the lane in the road network model in the previous round.

[0183] S5. Map the predicted road network position of the blind spot vehicle to the GPS coordinate form, update the blind spot vehicle coordinate point, and output the latest vehicle position.

[0184] Furthermore, if Fig. 9 As shown, step S5 includes:

[0185] S51. Compare the predicted road network position of the blind spot vehicle with the accumulated distances of each node on the lane centerline from the starting point to find the index of the line segment where the blind spot vehicle is located. i .

[0186] S52, according to the determined line segment index i , get the line segment from the list of preset polyline shapes for lane centerlines iThe starting and ending point coordinates of the vehicle in the blind spot are calculated through the conversion formula.

[0187] S53: Using the calculated GPS coordinates of the blind spot vehicle as the vehicle's latest current position, updating the coordinate point information of the blind spot vehicle and outputting it for subsequent use or display.

[0188] In yet another specific embodiment, based on the vehicle road network position , the cumulative distance from each node on the lane centerline to the starting point By contrast, use a fast search algorithm such as binary search to find the index of the line segment. i ,satisfy: .

[0189] Get line segments from a polyline shape i Starting and ending points, the vehicle GPS coordinates can be obtained through the conversion formula :

[0190] The conversion formula is:

[0191] ;

[0192] In the formula, Predict the road network position of vehicles in the blind spot, is the coordinate of the starting point of the line segment where the vehicle in the blind spot is located, is the coordinate of the end point of the line segment where the vehicle in the blind spot is located, The predicted road network position of the vehicle in the blind spot during the current period. , They are respectively the distances from the starting point and end point of the line segment where the vehicle in the blind spot is located to the starting point of the lane centerline of the line segment.

[0193] In addition, an embodiment of the present invention provides a vehicle GPS trajectory prediction system based on road network location information, comprising:

[0194] The model forming module is used to extract model data from the acquired map data and process it to form a road network model.

[0195] The trajectory judgment module is used to receive and track the real-time GPS coordinate data of the detected vehicle and judge whether the vehicle trajectory is interrupted according to the timestamp.

[0196] The position projection module is used to convert the historical GPS coordinates of blind spot vehicles into projection positions in the road network model through step-by-step retrieval and projection calculation if the vehicle trajectory is interrupted and the position of the blind spot vehicle is not updated and the latest GPS coordinate data is not transmitted.

[0197] The position prediction module is used to calculate the speed of the blind spot vehicle in the previous period based on the following model and the projection position, search for the front vehicle and calculate the speed difference and position difference between the front and rear vehicles, update the acceleration and speed of the blind spot vehicle, and determine the road network predicted position of the blind spot vehicle.

[0198] The position update module is used to map the road network predicted position of the blind spot vehicle into GPS coordinate form, update the blind spot vehicle coordinate point, and output the latest vehicle position.

[0199] Furthermore, an embodiment of the present invention provides a vehicle GPS trajectory prediction device based on road network location information, comprising: at least one database; and a memory communicatively connected to the at least one database; wherein the memory stores instructions that can be executed by the at least one database, and the instructions are executed by the at least one database so that the at least one database can execute the vehicle GPS trajectory prediction method based on road network location information as described above.

[0200] When these instructions are executed by the database, they will guide the database to perform a series of complex operations, thereby realizing the vehicle GPS trajectory prediction method based on road network location information. In simple terms, through this set of equipment and the instructions stored in it, we can use road network location information to predict the vehicle's GPS trajectory.

[0201] The emergence of this kind of equipment will undoubtedly greatly improve the accuracy and efficiency of vehicle trajectory prediction, and will have a profound impact on many fields such as intelligent transportation systems, vehicle scheduling, and path planning. By analyzing and processing a large amount of road network location information in real time, the device can provide accurate and timely vehicle trajectory prediction data for various applications, helping to upgrade the intelligence of related industries.

[0202] Furthermore, an embodiment of the present invention provides a computer-readable medium having computer executable instructions stored thereon, and when the executable instructions are executed by a processor, the vehicle GPS trajectory prediction method based on road network location information as described above is implemented.

[0203] In other words, this computer-readable medium is not only a carrier for data storage, but also the key to realizing the vehicle GPS trajectory prediction function. Through it, users can easily deploy the prediction method on various computing devices and use the computing power of the processor to quickly and accurately complete the vehicle trajectory prediction task. This is of great significance for improving the performance of intelligent transportation systems, optimizing vehicle scheduling strategies, and improving path planning solutions.

[0204] In summary, the embodiments of the present invention not only provide a vehicle GPS trajectory prediction method based on road network location information, but also construct corresponding systems, devices and media to fully support the implementation of this prediction method. Fig.10As shown in the figure, the entire prediction process starts with the acquisition and loading of road network data and vehicle trajectory GPS coordinate data, and then goes through a series of refined processing steps, such as selecting vehicles whose positions have not been updated, determining the update status of GPS coordinate points, converting GPS point coordinates to road network positions, etc., and then combining the car-following model to update vehicle acceleration, speed and position, and finally converting the vehicle road network position into the latest GPS coordinates, and completing the vehicle position update. This process is repeated until the positions of all vehicles are updated.

[0205] The core purpose of the present invention is to overcome the limitations of the prior art and strive to provide a more accurate, comprehensive and adaptable vehicle GPS trajectory prediction solution. In view of the problem that the simple linear or polynomial fitting methods in the prior art cannot accurately capture the nonlinear characteristics of vehicle motion, the present invention innovatively introduces a car-following model, which aims to accurately reflect the real motion state of the vehicle in complex traffic scenarios. By fully considering the randomness, mutation and high nonlinearity of traffic flow, the prediction accuracy and reliability of the present invention are significantly improved.

[0206] At the same time, the car-following model introduced by the present invention is not only computationally efficient, but also can save system resources, thereby providing timely prediction results, which is particularly important in real-time traffic management and navigation. In addition, the present invention also attaches great importance to geographical location information such as roads and lanes and status information of surrounding vehicles, and incorporates these key elements into the prediction model to achieve more comprehensive and accurate vehicle trajectory prediction.

[0207] It is worth mentioning that the present invention performs well in dealing with situations where data is temporarily missing. By effectively utilizing multi-dimensional information such as vehicle location history data, road information, and surrounding vehicle status, the present invention can accurately restore and predict the vehicle's GPS trajectory, thereby effectively reducing errors. This means that in complex and changing traffic environments, whether on normal roads or in special roads such as tunnels and bridges, the present invention can provide accurate, timely, and effective support for vehicle navigation and traffic management.

[0208] In general, the advantages of the present invention are mainly reflected in the following aspects: first, the prediction accuracy is greatly improved, thanks to the use of more advanced prediction models and algorithms that adapt to nonlinear characteristics; second, the information fusion is more comprehensive, and by fully integrating various related data, a richer and more comprehensive basis is provided for prediction; third, the ability to cope with data missing is strong, which ensures the stability of the system and the continuity of the trajectory.

[0209] Since the system / device described in the above embodiments of the present invention is a system / device used to implement the method of the above embodiments of the present invention, a person skilled in the art can understand the specific structure and deformation of the system / device based on the method described in the above embodiments of the present invention, and thus will not be described in detail here. All systems / devices used in the method of the above embodiments of the present invention belong to the scope of protection of the present invention.

[0210] It should be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0211] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions.

[0212] It should be noted that the word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention may be implemented by means of hardware comprising a number of different components and by means of a suitably programmed computer. Among the several means listed, several of these means may be embodied by the same hardware. The use of the words first, second, third, etc., is for convenience of expression only and does not indicate any order. These words may be understood as part of the name of the component.

[0213] In addition, it should be noted that, in the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0214] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments after knowing the basic creative concept. Therefore, the technical solution should be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0215] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the technical solution of the present invention and its equivalent technology, the present invention should also include these modifications and variations.

Claims

1. A vehicle GPS trajectory prediction method based on road network location information, characterized in that: include: Extracting model data from the acquired map data and processing it to form a road network model; Receive and track the real-time GPS coordinate data of the detected vehicle, and determine whether the vehicle trajectory is interrupted based on the timestamp; If the vehicle trajectory is interrupted, for the blind spot vehicles whose positions are not updated and the latest GPS coordinate data is not transmitted, the historical GPS coordinates of the blind spot vehicles are converted into the projected positions in the road network model through step-by-step retrieval and projection calculation; The speed of the vehicle in the blind spot in the previous period is calculated based on the following model and the projected position, the front vehicle is searched and the speed difference and position difference between the front and rear vehicles are calculated, and the acceleration and speed of the vehicle in the blind spot are updated to determine the predicted road network position of the vehicle in the blind spot; Map the predicted road network position of the blind spot vehicle to GPS coordinates, update the blind spot vehicle coordinates, and output the latest vehicle position; Among them, the speed of the blind spot vehicle in the previous period is calculated according to the following model and the projection position, the front vehicle is searched and the speed difference and position difference between the front and rear vehicles are calculated, and the acceleration and speed of the blind spot vehicle are updated to determine the road network predicted position of the blind spot vehicle. Calculate the speed of the vehicle in the blind spot during the last set period of time according to the distance traveled by the vehicle in the blind spot during the last set period of time; According to the projection position of the blind spot vehicle in the road network model, in the same lane, all other vehicles whose cumulative distance from the starting point is greater than the projection position of the blind spot vehicle in the road network model are screened out, and the vehicle closest to the blind spot vehicle is selected and marked as the front vehicle; Calculate the speed difference and position difference between the blind spot vehicle and the vehicle in front; Apply the intelligent driving formula in the car-following model, combine the speed difference, position difference, vehicle expected speed, maximum acceleration, comfortable deceleration, acceleration coefficient, minimum safety distance and safe headway time to calculate the expected safe distance and acceleration of vehicles in the blind spot; The speed of the current period is obtained based on the calculated acceleration, and the speed of the vehicle in the blind spot during the previous set period is updated to obtain the distance traveled during the current period. Determine the predicted road network position of the blind spot vehicle in the current period based on the projection position of the blind spot vehicle on the lane in the previous period and the distance it has traveled in the current period; in, The intelligent driving formula is: ； In the formula, is the expected safe distance between vehicles, v n The speed of the vehicle in the blind spot during the last set period of time The speed difference between the vehicle in the blind spot and the vehicle in front, S 0 is the minimum safety distance, T For safe headway, a is its maximum acceleration, b For comfortable deceleration, a n is the acceleration, v 0 is the desired speed of the vehicle, is the acceleration coefficient, It is the position difference between the vehicle in the blind spot and the vehicle in front.

2. The vehicle GPS trajectory prediction method based on road network location information as claimed in claim 1, characterized in that: Extracting model data from the acquired map data and processing it to form a road network model includes: Identify and extract lane centerline data including a lane centerline number, a lane number, and a set of longitude and latitude coordinate points from a starting point to an end point from the acquired map data; Identify and extract lane geographic shape information including lane number, road section number, and longitude and latitude point set from the acquired map data; Identify and extract the road segment geographic shape information including the road segment number, the road number to which it belongs, and the set of longitude and latitude coordinate points from the acquired map data; Identify and extract road geographic shape information including road numbers and longitude and latitude coordinate point sets from the acquired map data; Standardize lane centerline data, lane surface data, section surface data, and road surface data, and record the upstream and downstream relationships and subordinate relationships between each data; A road network model is formed based on the standardized data and the upstream, downstream and subordinate relationships between the various data.

3. The vehicle GPS trajectory prediction method based on road network location information as claimed in claim 1, characterized in that: Receiving and tracking the real-time GPS coordinate data of the detected vehicle and judging whether the vehicle track is interrupted according to the timestamp includes: Receive real-time GPS coordinate data of the vehicle detected by the detection equipment; Using the target ID as the unique identifier of each vehicle, the vehicle's location history data is tracked at fixed time intervals and integrated into a set of historical location points for each vehicle, where each location point contains the vehicle's latitude and longitude information and the timestamp of the corresponding point; Extract the latest location point data from the historical location point set of each vehicle; Compare the timestamp of the latest location point data with the current time and calculate the time difference; If the calculated time difference is greater than the preset threshold, the current vehicle trajectory is determined to be interrupted, indicating that the vehicle position information at the current time is missing, and the next processing flow is entered; If the calculated time difference is less than or equal to the preset threshold, it is determined that the vehicle detection trajectory is not interrupted and the latest position of the vehicle is directly output.

4. The vehicle GPS trajectory prediction method based on road network location information as claimed in claim 2, characterized in that: If the vehicle trajectory is interrupted, for the blind spot vehicle whose position is not updated and no latest GPS coordinate data is transmitted, the historical GPS coordinates of the blind spot vehicle are converted into the projection position in the road network model through step-by-step retrieval and projection calculation, including: Identify and select vehicles whose positions have not been updated and whose latest GPS coordinate data has not been transmitted as blind spot vehicles that need to be processed; Using the historical GPS coordinates of the vehicle in the blind spot and combining it with the geographical shape information of the road, the road to which the historical GPS coordinates of the vehicle in the blind spot belong is retrieved in the road network model, and the corresponding road identifier is recorded; According to the road identifier, all the road sections included in the corresponding road are screened out from the preset road section information table, and the road section where the blind spot vehicle coordinate point is located is retrieved in the road network model in combination with the road section geographic shape information, and the corresponding road section identifier is recorded; According to the road segment identifier, all lanes included in the corresponding road segment are screened from the preset lane information table, and the lane where the coordinate point of the blind spot vehicle is located is retrieved from the road network model in combination with the lane geographic shape information, and the corresponding lane identifier is recorded; According to the lane identifier, extracting the corresponding lane centerline represented in the form of a series of broken line segments from a preset lane centerline information table; For each coordinate point of a vehicle in the blind spot, the following operations are repeated until all line segments of the lane centerline are processed: a line segment is selected from a preset polyline shape list of the lane centerline, and based on the projection relationship between the coordinate point of the vehicle in the blind spot and the selected line segment, whether the projection point is inside the line segment is determined, and if it is inside the line segment, the line segment is marked; Among all the marked line segments, select a line segment that minimizes the distance between the blind spot vehicle coordinate point and the line segment as the projection target line segment; According to the index of the projected target line segment, the accumulated distance from the starting point of the line segment to the starting point of the lane centerline is obtained, so as to determine the projection position of the blind spot vehicle in the road network model.

5. The vehicle GPS trajectory prediction method based on road network location information as claimed in claim 1, characterized in that: Select a line segment from the preset polyline shape list of the lane centerline, and determine whether the projection point is inside the line segment based on the projection relationship between the blind spot vehicle coordinate point and the selected line segment. If it is inside the line segment, mark the line segment including: Select a line segment AB from the preset polyline shape list of the lane centerline as the current processing object; Determine the coordinates of the starting point A and the end point B of the current processing object and the coordinate point P of the blind spot vehicle, and obtain the first vector from the starting point A of the line segment to the coordinate point P of the blind spot vehicle , and a second vector from the starting point A to the end point B of the line segment ; According to the first vector and the second vector The angle between them is used to find the first vector In the second vector The projection length on the second vector is divided by the projection length The scalar length of the coordinate point P is obtained by The projection relationship; Based on the projection relationship, determine whether the projection point of the coordinate point P of the blind spot vehicle is within the selected line segment AB; If the value of the projection relationship is within the set interval, it is determined that the projection point of the coordinate point P is within the line segment AB, and the line segment is marked; If the value of the projection relationship is not within the set interval, it is determined that the projection point of the coordinate point P is not within the line segment AB, and the line segment is skipped and the next line segment is found as the current processing object for processing.

6. The vehicle GPS trajectory prediction method based on road network location information as claimed in claim 1, characterized in that: Map the predicted road network position of the blind spot vehicle to the GPS coordinate form, and update the blind spot vehicle coordinate point. Output the latest vehicle position including: Compare the predicted road network position of the blind spot vehicle with the accumulated distance from each node of the lane centerline to the starting point to find the index of the line segment where the blind spot vehicle is located i ; According to the determined line segment index i , get the line segment from the list of preset polyline shapes for lane centerlines i The starting and ending coordinates of the vehicle are calculated by the conversion formula to obtain the GPS coordinates of the vehicle in the blind spot. The calculated GPS coordinates of the blind spot vehicle are used as the vehicle's latest current position, and the coordinate point information of the blind spot vehicle is updated and output for subsequent use or display; in, The conversion formula is: ； In the formula, Predict the road network position of vehicles in the blind spot, is the coordinate of the starting point of the line segment where the vehicle in the blind spot is located, is the coordinate of the end point of the line segment where the vehicle in the blind spot is located, The predicted road network position of the vehicle in the blind spot during the current period. , They are respectively the distances from the starting point and end point of the line segment where the vehicle in the blind spot is located to the starting point of the lane centerline of the line segment.

7. A vehicle GPS trajectory prediction system based on road network location information, characterized in that: include: A model forming module, used for extracting model data from the acquired map data and processing it to form a road network model; The track judgment module is used to receive and track the real-time GPS coordinate data of the detected vehicle and judge whether the vehicle track is interrupted according to the timestamp; The position projection module is used to convert the historical GPS coordinates of the blind spot vehicle into the projection position in the road network model through step-by-step retrieval and projection calculation if the vehicle trajectory is interrupted and the position of the blind spot vehicle is not updated and the latest GPS coordinate data is not transmitted; The position prediction module is used to calculate the speed of the blind spot vehicle in the previous period based on the following model and the projected position, search for the leading vehicle and calculate the speed difference and position difference between the leading and trailing vehicles, and update the acceleration and speed of the blind spot vehicle to determine the road network predicted position of the blind spot vehicle; The position update module is used to map the road network predicted position of the blind spot vehicle into the GPS coordinate form, update the blind spot vehicle coordinate point, and output the latest vehicle position; Among them, the speed of the blind spot vehicle in the previous period is calculated according to the following model and the projection position, the front vehicle is searched and the speed difference and position difference between the front and rear vehicles are calculated, and the acceleration and speed of the blind spot vehicle are updated to determine the road network predicted position of the blind spot vehicle. Calculate the speed of the vehicle in the blind spot during the last set period of time according to the distance traveled by the vehicle in the blind spot during the last set period of time; According to the projection position of the blind spot vehicle in the road network model, in the same lane, all other vehicles whose cumulative distance from the starting point is greater than the projection position of the blind spot vehicle in the road network model are screened out, and the vehicle closest to the blind spot vehicle is selected and marked as the front vehicle; Calculate the speed difference and position difference between the blind spot vehicle and the vehicle in front; Apply the intelligent driving formula in the car-following model, combine the speed difference, position difference, vehicle expected speed, maximum acceleration, comfortable deceleration, acceleration coefficient, minimum safety distance and safe headway time to calculate the expected safe distance and acceleration of vehicles in the blind spot; The speed of the current period is obtained based on the calculated acceleration, and the speed of the vehicle in the blind spot during the previous set period is updated to obtain the distance traveled during the current period. Determine the predicted road network position of the blind spot vehicle in the current period based on the projection position of the blind spot vehicle on the lane in the previous period and the distance it has traveled in the current period; in, The intelligent driving formula is: ； In the formula, is the expected safe distance between vehicles, v n The speed of the vehicle in the blind spot during the last set period of time The speed difference between the vehicle in the blind spot and the vehicle in front, S 0 is the minimum safety distance, T For safe headway, a is its maximum acceleration, b For comfortable deceleration, a n is the acceleration, v 0 is the desired speed of the vehicle, is the acceleration coefficient, It is the position difference between the vehicle in the blind spot and the vehicle in front.

8. A vehicle GPS trajectory prediction device based on road network location information, characterized in that: include: at least one database; and a memory in communication with at least one database; Wherein, the memory stores instructions that can be executed by at least one database, and the instructions are executed by at least one database so that at least one database can execute the vehicle GPS trajectory prediction method based on road network location information as described in any one of claims 1-6.

9. A computer-readable medium having computer-executable instructions stored thereon, characterized in that: When the executable instructions are executed by the processor, the vehicle GPS trajectory prediction method based on road network location information as described in any one of claims 1 to 6 is implemented.

Images