Vehicle trajectory prediction method, device, equipment and program product

Abstract

The invention relates to the field of intelligent driving, in particular to a vehicle track prediction method, device and equipment and a program product. The method comprises the following steps: acquiring a lane set related to a planned path of a vehicle; related vehicles of the own vehicle are searched in the lane set, and vehicle states of the own vehicle and the related vehicles are obtained; according to the vehicle state of the vehicle, deducing a preliminary prediction track of the vehicle within a preset duration, and according to the vehicle state of the related vehicle, deducing a preliminary prediction track of the related vehicle within the preset duration; collision detection is carried out according to the preliminary prediction trajectories of the related vehicle and the own vehicle in the preset duration to generate a collision detection result; and determining a vehicle track prediction result of the vehicle according to the collision detection result. According to the method, on the basis of ensuring the prediction precision, the consumption of computing resources can be effectively reduced, the influence of map precision is avoided, and the adaptability of an actual environment is improved.

Description

Technical Field

[0001] This application relates to the field of intelligent driving, and in particular to a vehicle trajectory prediction method, device, equipment, and program product. Background Technology

[0002] In current research on intelligent transportation and autonomous driving, vehicle trajectory prediction is one of the key technologies. Existing trajectory prediction methods include prediction based on the vehicle's own operating state, or using fully interactive prediction to comprehensively and interactively model the trajectories of all surrounding vehicles, or trajectory prediction based on fixed route information provided by high-precision maps.

[0003] In real-world deployment scenarios, especially in enclosed industrial parks, high-precision map routes may differ from actual road conditions, such as temporary road changes or obstacle occlusion, affecting the accuracy of trajectory prediction based on high-precision maps and hindering environmental adaptability. When predicting based on the vehicle's own operating status, the dynamic impact of surrounding vehicles on the vehicle leads to inaccurate prediction results. Furthermore, using fully interactive prediction computation is computationally complex, and comprehensive interactive modeling would result in significant resource waste. Summary of the Invention

[0004] In view of this, embodiments of this application provide a vehicle trajectory prediction method, apparatus, device, and program product to solve the problems of inaccurate prediction results, which lead to a lot of resource waste and are not conducive to improving environmental adaptability when performing vehicle trajectory prediction in the prior art.

[0005] A first aspect of this application provides a vehicle trajectory prediction method, the method comprising:

[0006] Obtain the set of lanes associated with the vehicle's planned path;

[0007] Search for related vehicles of the vehicle in the lane set, and obtain the vehicle status of the vehicle and the related vehicles;

[0008] Based on the vehicle status of the self-vehicle, a preliminary predicted trajectory of the self-vehicle within a preset time period is deduced; and based on the vehicle status of the related vehicles, a preliminary predicted trajectory of the related vehicles within a preset time period is deduced.

[0009] Collision detection results are generated based on the preliminary predicted trajectories of the relevant vehicles and the autonomous vehicle within a preset time period.

[0010] The vehicle trajectory prediction result is determined based on the collision detection result.

[0011] In conjunction with the first aspect, in a first possible implementation of the first aspect, the vehicle trajectory prediction result of the vehicle is determined based on the collision detection result;

[0012] When the collision detection result indicates that at least one of the vehicle and all related vehicles is at risk of collision within the predicted time period, the preliminary predicted trajectories of the related vehicles and the vehicle within the preset time period are input into a preset intelligent driver model to update the vehicle trajectory prediction result of the vehicle.

[0013] When the collision detection result indicates that the vehicle has no risk of collision with any related vehicles within a preset time period, the preliminary predicted trajectory of the vehicle within the preset time period is determined as the trajectory prediction result of the vehicle.

[0014] In conjunction with the first aspect, in the second possible implementation of the first aspect, the set of lanes related to the planned path of the vehicle is obtained, including:

[0015] Obtain the lane currently occupied by the vehicle;

[0016] In addition, the lanes involved in the planned path of the vehicle are obtained.

[0017] In conjunction with the first aspect, in a third possible implementation of the first aspect, searching for the relevant vehicle of the self-driving vehicle in the set of lanes includes:

[0018] Get the current driving lanes of vehicles surrounding your vehicle;

[0019] When the current driving lane intersects with a lane in the lane set, the vehicle is determined to belong to the related vehicles of the vehicle.

[0020] In conjunction with the first aspect, in a fourth possible implementation of the first aspect, searching for the relevant vehicle of the self-driving vehicle in the set of lanes includes:

[0021] The vehicle obtains the speed and direction of movement of surrounding vehicles, predicts the endpoint position after a preset time, and determines the movement line segment based on the vehicle's current position and the endpoint position.

[0022] The polygon where the vehicle is located is generated based on the boundary lines of the lanes traversed by the vehicle within the preset time period.

[0023] When the moving line segment intersects with the polygon, it is determined that the vehicle belongs to the relevant vehicle of the vehicle.

[0024] In conjunction with the first aspect, in the fifth possible implementation of the first aspect, the preliminary predicted trajectory of the vehicle within a preset time period is deduced based on the vehicle's state, including:

[0025] When the vehicle's state is to maintain its position, the preliminary predicted trajectory of the vehicle within a preset time period is deduced to be a constant speed trajectory or a constant acceleration trajectory along the center line of the current lane.

[0026] When the vehicle's status is a lane change intention, the preliminary predicted trajectory of the vehicle within a preset time period is a transition trajectory from the current lane to the target lane.

[0027] In conjunction with the first aspect, in the sixth possible implementation of the first aspect, the preliminary predicted trajectory of the relevant vehicle within a preset time period is deduced based on the vehicle status of the relevant vehicle, including:

[0028] If the distance between the position of the relevant vehicle and the center of its lane is less than a first preset distance, and the angle between the speed direction of the relevant vehicle and the lane line of its lane is less than a first preset angle, then the relevant vehicle is determined to have a holding intention, and the holding trajectory of the relevant vehicle is deduced based on the distance of the relevant vehicle from the center of the lane.

[0029] If the angle between the speed direction of the relevant vehicle and the lane line is greater than a first preset angle and less than a second preset angle, then the relevant vehicle is determined to have the intention to change lanes, and the lane change trajectory of the relevant vehicle is deduced based on the center line of the current lane and the target lane of the lane change.

[0030] If the angle between the speed direction of the relevant vehicle and the lane line is greater than a second preset angle and less than a third preset angle, then it is determined that the relevant vehicle intends to cross the lane, and the crossing trajectory of the relevant vehicle is deduced based on the movement trend of the relevant vehicle.

[0031] If the angle between the speed direction of the related vehicle and the speed direction of the vehicle is greater than a third preset angle, then the related vehicle is determined to be traveling in the opposite direction, and the trajectory of the related vehicle is deduced based on the opposite direction of the center line of the lane where the related vehicle is located. The lane line direction of the vehicle is the same as the driving direction of the vehicle.

[0032] A second aspect of this application provides a vehicle trajectory prediction device, the device comprising:

[0033] The lane set acquisition unit is used to acquire the lane set related to the vehicle's planned path;

[0034] The related vehicle search unit is used to search for related vehicles of the vehicle in the lane set and obtain the vehicle status of the vehicle and the related vehicles;

[0035] The preliminary trajectory prediction unit is used to predict the preliminary trajectory of the vehicle within a preset time period based on the vehicle status of the vehicle, and to predict the preliminary trajectory of the related vehicles within a preset time period based on the vehicle status of the related vehicles.

[0036] The collision detection unit is used to perform collision detection and generate collision detection results based on the preliminary predicted trajectories of the relevant vehicles and the self-vehicle within a preset time period.

[0037] The vehicle trajectory prediction result determination unit is used to determine the vehicle trajectory prediction result of the vehicle based on the collision detection result.

[0038] A third aspect of this application provides a vehicle trajectory prediction device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the vehicle trajectory prediction device performs the method as described in any of the first aspects.

[0039] A fourth aspect of this application provides a computer program product that, when run on a computer, causes the computer to execute the methods described in the first aspect or its various implementations.

[0040] A fifth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in any of the first aspects.

[0041] A sixth aspect of this application provides a chip for implementing the methods in the various implementations of the first aspect described above. Specifically, the chip includes a processor for calling and running a computer program from a memory, causing a device equipped with the chip to perform the methods as described in the first aspect or its various implementations.

[0042] The beneficial effects of this application embodiment compared with the prior art are as follows: This application embodiment obtains a set of lanes related to the planned path of the vehicle, searches for relevant vehicles in the set of lanes, infers the preliminary predicted trajectory of the vehicle within a preset time period based on the vehicle status of the vehicle, infers the preliminary predicted trajectory of the relevant vehicles within a preset time period based on the vehicle status of the relevant vehicles, performs collision detection based on the preliminary predicted trajectory, and determines the vehicle trajectory prediction result based on the collision detection result. Since this method determines the relevant set of lanes based on the planned path of the vehicle, lanes can be filtered, and relevant vehicles can be determined based on the filtered set of lanes. Only vehicles affecting the vehicle are subject to trajectory prediction, which can effectively reduce the number of vehicles that need to be predicted, thereby reducing the computational resources required. Furthermore, since this method predicts the trajectory of relevant vehicles filtered from the set of lanes, it is not affected by the accuracy of the map, which can effectively improve the adaptability to the actual environment. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram illustrating the implementation process of a vehicle trajectory prediction method provided in an embodiment of this application;

[0045] Figure 2 This is a schematic diagram illustrating the implementation process of a method for determining the relevant vehicles of a self-driving vehicle, provided in an embodiment of this application.

[0046] Figure 3 This is a schematic diagram illustrating the identification of an intruding vehicle according to an embodiment of this application;

[0047] Figure 4 This is a schematic diagram of the preliminary predicted trajectory of a vehicle and related vehicles provided in an embodiment of this application;

[0048] Figure 5 This is another schematic diagram of the preliminary predicted trajectory of a vehicle and related vehicles provided in the embodiments of this application;

[0049] Figure 6 This is another schematic diagram of the preliminary predicted trajectory of a vehicle and related vehicles provided in the embodiments of this application;

[0050] Figure 7 This is a schematic diagram of a vehicle trajectory prediction device provided in an embodiment of this application;

[0051] Figure 8This is a schematic diagram of a vehicle trajectory prediction device provided in an embodiment of this application. Detailed Implementation

[0052] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0053] To illustrate the technical solution described in this application, specific embodiments are provided below.

[0054] In the current field of intelligent transportation and autonomous driving, vehicle trajectory prediction is one of the key technologies supporting driving decisions. Existing methods mainly fall into three categories: first, prediction relies solely on the vehicle's own operating state, ignoring the dynamic influence of surrounding vehicles; second, fully interactive prediction is used to comprehensively and interactively model the trajectories of all surrounding vehicles; and third, trajectory extrapolation is based on fixed route information provided by high-precision maps.

[0055] However, the above methods face significant challenges in practical applications, especially in closed environments such as industrial parks. First, high-precision maps often fail to reflect dynamic road conditions such as temporary road changes or obstacle occlusion in real time, causing map-based predictions to deviate from reality. Second, prediction methods based solely on the vehicle's own state have limited accuracy because they do not consider the interactive effects of other vehicles' behavior on the vehicle. While fully interactive prediction can characterize the interactions between vehicles, its high computational complexity can lead to resource waste. Furthermore, limited by the vehicle's perception field of view, fully interactive modeling may result in "over-prediction" or "under-prediction" of trajectories, i.e., over-responding to irrelevant vehicles or under-judging potential risks.

[0056] To address the aforementioned issues, this application proposes a vehicle trajectory prediction method. This method focuses on lane information within the planned path of the vehicle, accurately models the dynamic interaction impact of other vehicles on the vehicle, and generates high-precision vehicle trajectory prediction results. This solves problems such as high computational load, inaccurate prediction when high-precision maps do not match actual road conditions, and unreasonable interactive prediction in existing technologies. It provides reliable decision support for intelligent driving systems and improves driving safety and traffic efficiency. Figure 1 The implementation flow of this method is illustrated in detail below:

[0057] In S101, obtain the set of lanes associated with the vehicle's planned path.

[0058] In this embodiment, the self-driving vehicle is the executing entity of the vehicle trajectory prediction method. The self-driving vehicle performs vehicle control based on the vehicle trajectory prediction result obtained by the vehicle trajectory prediction method.

[0059] This planned route can be the route planned continuously for a preset duration starting from the current time. The planned route can obtain the overall planned route through the vehicle navigation planning module, and determine local planned routes based on the vehicle's surrounding environmental information, including surrounding vehicle information and road condition information, to adjust the overall planned route.

[0060] The set of lanes related to the planned path is the set of lanes relevant to the vehicle's journey. This set of lanes can include the vehicle's current lane and potential target lanes along the planned path, including lane-changing lanes and lanes that may be used for other lanes.

[0061] For example, if the planned path of the vehicle is to turn left at the next intersection, then the extracted lane set includes: 1. the straight lane where the vehicle is currently located; 2. the left-turn lane that needs to be changed into ahead; 3. the lane to enter after turning left at the intersection.

[0062] Alternatively, the lane set for the vehicle can be determined based on the driving intention. For example, if the driving intention is to maintain the lane, the lane set is the current lane. If the driving intention is to change lanes, the lane set is the current vehicle and the target lane for the lane change.

[0063] Once the lane set is determined, lane information within the set can be obtained, including lane center information, lane boundaries, lane width, and lane type. Lane boundaries include the left and right boundaries. Lane type information includes types such as main road and ramp.

[0064] In S102, the relevant vehicles of the vehicle are searched in the lane set, and the vehicle status of the vehicle and the relevant vehicles is obtained.

[0065] After determining the lane set, the current lanes of vehicles in front of and to the sides of the vehicle can be obtained. If there is an intersection between the current lane sets of surrounding vehicles, then the vehicles with the intersection are determined to be related to the vehicle.

[0066] For example, when a vehicle travels along lane L2 to the next intersection, among the surrounding vehicles detected, the first vehicle is traveling in lane L2 directly in front of the vehicle. Since the lane L2 where the first vehicle is located is the same lane as lane L2 in the vehicle's lane set, it is confirmed that the first vehicle intersects with the vehicles in the lane set (the two are in the same lane), and the first vehicle belongs to the vehicle's related vehicles.

[0067] At this point, assuming that the second vehicle of the surrounding vehicles is traveling parallel to the vehicle in the adjacent lane L1, since lane L1 has no intersection with any lane in the lane set, i.e., lane L2, the second vehicle will not be marked as a relevant vehicle. This filters out the second vehicle that is irrelevant to the vehicle's travel, avoids the calculation of vehicles in irrelevant lanes, and saves computing resources.

[0068] Assuming the vehicle's planned path is "currently in L2, turns to L1 after 100 meters, then goes straight," the lane set is: [Lane L2, Lane L1]. Assuming that among the surrounding vehicles, the first vehicle is traveling in lane L1 and the second vehicle is traveling in lane L2, since lane L1 of the first vehicle intersects with lane L1 in the lane set, and lane L2 of the second vehicle intersects with lane L2 in the lane set, the identified relevant vehicles are: Vehicle 1 and Vehicle 2.

[0069] In determining relevant vehicles, this application embodiment, in addition to determining whether the current driving lane of surrounding vehicles intersects with the lane set, may also include dynamically predicting intruding vehicles. Determining relevant vehicles includes vehicles intruding into the planned path, specifically as follows: Figure 2 As shown, it includes:

[0070] In S201, the speed and direction of movement of surrounding vehicles are obtained, the endpoint position after a preset time is predicted, and the movement line segment is determined based on the current position of the vehicle and the endpoint position.

[0071] The vehicle uses sensors, including cameras or radar, to detect the movement information of surrounding vehicles, including their speed and direction of movement. Based on the speed and direction of movement, the movement trend of surrounding vehicles can be predicted within a preset time period.

[0072] For example, if the primary vehicle is traveling in a straight line in lane L1, and the second vehicle is traveling in lane L2 adjacent to lane L1, the lane L2 where the second vehicle is currently located does not intersect with the lane set of the primary vehicle's planned path, so the second vehicle cannot be classified as a related vehicle. However, if the motion information of the second vehicle is detected, with a speed of 2.8 ms and a speed direction of left front, after a preset time, such as 6 seconds, the distance traveled would be 6s * 2.8m / s = 16.8m. Based on the speed direction and the traveled distance, the endpoint position P1 can be determined. Combined with the first vehicle's current position P0, the motion line segment P0P1 of the first vehicle can be obtained.

[0073] In S202, the polygon where the vehicle is located is generated based on the boundary lines of the lanes traversed by the vehicle within the preset time period.

[0074] Within the preset time period, while the vehicle is maintaining its lane, its final position can be determined based on its speed. By combining the vehicle's starting and ending positions with the lane lines of the lane it traverses, a polygon representing the vehicle can be generated from the lane boundaries traversed within the preset time period. When the lane is a straight line, the vehicle's polygon can be rectangular.

[0075] If the vehicle changes lanes within a preset time period, the polygon generated by the vehicle will include lane lines of different lanes.

[0076] In S203, when the moving line segment intersects with the polygon, it is determined that the vehicle belongs to the vehicle's related vehicles.

[0077] When a moving line segment intersects with a polygon, it indicates that surrounding vehicles are encroaching on the planned path of the vehicle, which may lead to a collision. Therefore, surrounding vehicles can be added to the vehicle's relevant vehicle list.

[0078] Figure 2 The method for detecting related vehicles, as shown, directly determines the polygon where the vehicle is located using lane lines, and determines the movement line segments using the current and endpoint positions of surrounding vehicles. This facilitates quick determination of whether a vehicle is related to the vehicle, improving the efficiency of related vehicle detection. Beyond this, it can also determine the movement space of the vehicle and surrounding vehicles based on their dimensions, and determine whether the surrounding vehicles are related vehicles based on whether their movement spaces intersect with the vehicle's. If the movement spaces of surrounding vehicles intersect with the vehicle's, the surrounding vehicle is considered a related vehicle; otherwise, it is considered an unrelated vehicle.

[0079] for example Figure 3 As shown, based on the current position P2 of the vehicle and the endpoint position P3 after a preset time of 6 seconds, and combined with the two lane lines of the lane where the vehicle is located, polygon ABCD is determined. Among the surrounding vehicles, the current position of the first vehicle is P0, and the position of the first vehicle after a preset time of 6 seconds is P1. Connecting P0 and P1 yields the motion line segment P0P1. Since the motion line segment P0P1 is related to polygon ABCD, the first vehicle is an intruding vehicle, satisfying the requirements of a related vehicle. Therefore, the first vehicle is determined to be a related vehicle of the vehicle.

[0080] In S103, the preliminary predicted trajectory of the self-vehicle within a preset time period is deduced based on the vehicle status of the self-vehicle, and the preliminary predicted trajectory of the related vehicles within a preset time period is deduced based on the vehicle status of the related vehicles.

[0081] The vehicle's status includes information such as its position, speed, acceleration, driving intention, and dimensions. Based on the vehicle's status, a preliminary predicted trajectory for the vehicle within a preset time period can be determined according to the driving intention and speed.

[0082] The preset duration can be any value less than 10 seconds, or any value between 3 and 10 seconds.

[0083] When the vehicle's driving intention is to maintain its position, the preliminary predicted trajectory within a preset time period is a constant speed or constant acceleration trajectory along the center of the current lane. When the vehicle's driving intention is to change lanes, the preliminary predicted trajectory within a preset time period is a transition trajectory from the current lane to the target lane.

[0084] When the vehicle's driving intention is to maintain its position, it will strictly follow the current lane. Trajectory prediction is divided into lateral and longitudinal control: Laterally, the lane centerline is used as a reference path, and a control algorithm (such as a pure tracking algorithm) ensures that the vehicle's direction is consistent with the lane direction; longitudinally, a constant speed or constant acceleration model is used. For example, if the vehicle's lane is straight and its speed is uniform, the initial predicted trajectory is a uniform linear motion along the lane centerline. Simplifying the model improves computational efficiency and allows for rapid generation of the initial predicted trajectory.

[0085] When the vehicle's intention is to change lanes, a smooth transition trajectory needs to be generated. The lateral trajectory connects the centerlines of the current lane and the target lane using a curve interpolation algorithm (such as a fifth-order polynomial) to ensure continuous lateral acceleration and avoid abrupt changes; the longitudinal trajectory usually assumes that the speed is basically stable. For example, when the vehicle changes from lane L1 to L2, the initial predicted trajectory is a smooth S-shaped curve, with the lateral position gradually shifting while the longitudinal speed remains stable.

[0086] When extrapolating the preliminary predicted trajectory of relevant vehicles within a preset time period based on their vehicle states, the preliminary predicted trajectory of relevant vehicles within the preset time period can be determined according to the driving intentions of the relevant vehicles. The driving intentions of the relevant vehicles may include at least one of the following: lane keeping intention, lane changing intention, crossing intention, and oncoming intention. The driving intentions of relevant vehicles can be determined based on their vehicle states, thereby determining the preliminary predicted trajectory of the relevant vehicles within the preset time period.

[0087] For example, if the distance between the relevant vehicle's position and the center of its lane is less than a first preset distance (the first preset distance can be any value between 40-60cm, adjustable according to the lane width) and the angle between the relevant vehicle's velocity direction and the lane line of the lane where the relevant vehicle is located (which can be the nearest lane line) is less than a first preset angle (the first preset angle can be any value between 10-20 degrees, such as 10 degrees, 15 degrees, or 20 degrees), then it is determined that the relevant vehicle intends to maintain its position, and the vehicle's trajectory is deduced based on the distance the relevant vehicle deviates from the lane center. For example... Figure 4 As shown, the driving intentions of both vehicle 001 and the related vehicle 002 ahead of vehicle 001 are to maintain their position. The preliminary predicted trajectories of vehicle 001 and the related vehicle 002 are obtained as follows: Figure 4 As shown, all are forward straight trajectories, and the length of the trajectory is determined by the speed of the vehicle and related vehicles.

[0088] If the angle between the speed direction of the relevant vehicle and the lane line where the relevant vehicle is located is greater than a first preset angle and less than a second preset angle (which can be any value between 70 and 80 degrees, such as 70, 75, or 80 degrees), then it is determined that the relevant vehicle intends to change lanes, and the lane-changing trajectory of the relevant vehicle is deduced based on the center lines of the current lane and the target lane of the lane change. For example... Figure 5 As shown, the driving intention of vehicle 001 is to maintain its position, and its initial predicted trajectory within a preset time period is a forward straight trajectory, the length of which is determined by the vehicle's speed. The driving intention of related vehicle 003 is to change lanes, and its initial predicted trajectory is a smooth lane change to the lane where the vehicle is located, the length of which is related to the speed of the related vehicle.

[0089] If the angle between the speed direction of the relevant vehicle and the lane line is greater than a second preset angle and less than a third preset angle (the third preset angle is any angle between 130 and 140 degrees, such as 130, 135, or 140 degrees), then it is determined that the relevant vehicle intends to cross the lane, and the crossing trajectory of the relevant vehicle is deduced based on the movement trend of the relevant vehicle. Figure 6 As shown, the driving intention of the vehicle is to maintain its position. The initial predicted trajectory of the vehicle within a preset time period is a forward straight trajectory, the length of which is determined by the vehicle's speed. The initial predicted trajectory of the related vehicle 004, which intends to cross the lane, is a trajectory that crosses the lane line, the length of which is determined by the crossing speed of the related vehicle 004.

[0090] If the angle between the speed direction of the related vehicle and the speed direction of the vehicle is greater than a third preset angle, then the related vehicle is determined to be traveling in the opposite direction. The oncoming trajectory of the related vehicle is then deduced based on the opposite direction of the centerline of the lane in which the related vehicle is located. The lane line direction of the vehicle is the same as the vehicle's travel direction. Figure 6 As shown, the driving intention of the related vehicle 005 is to drive in the opposite direction. The driving direction of the related vehicle is opposite to that of the self-vehicle, and the angle between the two is 180 degrees, which is greater than the third preset angle. The preliminary predicted trajectory of the related vehicle 005 is determined to be the trajectory extending in the opposite direction of the self-vehicle, starting from the current position of the related vehicle. The length of the trajectory is determined according to the speed of the related vehicle 005.

[0091] In S104, collision detection results are generated based on the preliminary predicted trajectories of the relevant vehicles and the self-driving vehicle within a preset time period.

[0092] When performing collision detection between the initial predicted trajectory of the vehicle and the initial predicted trajectories of related vehicles, it is possible to first detect whether the initial predicted trajectory of the vehicle overlaps with the initial predicted trajectories of all related vehicles, or determine whether the closest distance between the vehicle and related vehicles is less than a second preset distance based on the initial predicted trajectory of the vehicle and the initial predicted trajectories of related vehicles. The second preset distance can be any value between 1.3 and 1.8 meters, such as 1.5 meters.

[0093] If the initial predicted trajectory of the vehicle overlaps with the initial predicted trajectories of all relevant vehicles, or if the initial predicted trajectory of the vehicle and the initial predicted trajectories of relevant vehicles determine that the closest distance between the vehicle and relevant vehicles is less than a second preset distance, then a collision risk is determined to exist. If the initial predicted trajectory of the vehicle does not overlap with the initial predicted trajectories of all relevant vehicles, and the initial predicted trajectory of the vehicle and the initial predicted trajectories of relevant vehicles determine that the closest distance between the vehicle and relevant vehicles is greater than or equal to the second preset distance, then no collision risk is determined to exist.

[0094] In S105, the vehicle trajectory prediction result of the vehicle is determined based on the collision detection result.

[0095] If no collision risk is detected between the vehicle and all related vehicles within a preset time period, the vehicle's preliminary predicted trajectory can be directly output as the vehicle trajectory prediction result.

[0096] If a collision risk is detected between the vehicle and at least one of the relevant vehicles within a preset time period, the vehicle's preliminary predicted trajectory and the preliminary preset trajectories of the relevant vehicles at risk of collision can be input into a preset intelligent driver model. The intelligent driver model then adjusts the vehicle's driving speed to avoid a collision with the relevant vehicles. Based on the adjusted driving speed, the vehicle's predicted trajectory is re-determined, resulting in the vehicle's trajectory prediction result.

[0097] In summary, the embodiments of this application have the following significant effects:

[0098] 1. Reduced computational load: To address the problem of excessive computation caused by trajectory-related factors in existing trajectory prediction, this application only considers relevant vehicles in the relevant lanes that affect the trajectory path of the vehicle for prediction, effectively reducing the amount of computation and improving the real-time performance and efficiency of the system.

[0099] 2. Targeted prediction: In scenarios such as closed industrial parks where the route on the precision map does not match the actual road conditions, the vehicle trajectory prediction method in this application can effectively filter out dynamic vehicles that affect the vehicle's path and perform targeted trajectory prediction to make effective decisions and enhance the system's adaptability in the actual environment.

[0100] 3. Adopting a reasonable interactive method: In view of the drawbacks of existing interactive or fully interactive trajectory prediction, the embodiments of this application adopt a semi-interactive method, that is, only making avoidance actions on the preliminary predicted trajectory with collision risk, simulating the deceleration of real rear vehicles, side vehicles, and oncoming vehicles, thereby reducing the computational complexity while ensuring prediction accuracy.

[0101] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0102] Figure 7 This is a schematic diagram of a vehicle trajectory prediction device provided in an embodiment of this application. The device includes:

[0103] Lane set acquisition unit 701 is used to acquire the lane set related to the planned path of the vehicle.

[0104] The related vehicle search unit 702 is used to search for related vehicles of the vehicle in the lane set and obtain the vehicle status of the vehicle and the related vehicles.

[0105] The preliminary trajectory prediction unit 703 is used to predict the preliminary trajectory of the vehicle within a preset time period based on the vehicle status of the vehicle, and to predict the preliminary trajectory of the related vehicles within a preset time period based on the vehicle status of the related vehicles.

[0106] The collision detection unit 704 is used to perform collision detection and generate collision detection results based on the preliminary predicted trajectories of the relevant vehicles and the self-vehicle within a preset time period.

[0107] The vehicle trajectory prediction result determination unit 705 is used to determine the vehicle trajectory prediction result of the vehicle based on the collision detection result.

[0108] Figure 7 The vehicle trajectory prediction device shown is, with Figure 1 The vehicle trajectory prediction method shown corresponds to this.

[0109] Figure 8 This is a schematic diagram of a vehicle trajectory prediction device provided in an embodiment of this application. Figure 8 As shown, the vehicle trajectory prediction device 8 of this embodiment includes: a processor 80, a memory 81, and a computer program 82, such as a vehicle trajectory prediction program, stored in the memory 81 and executable on the processor 80. When the processor 80 executes the computer program 82, it implements the steps in the various vehicle trajectory prediction method embodiments described above. Alternatively, when the processor 80 executes the computer program 82, it implements the functions of each module / unit in the various device embodiments described above.

[0110] For example, the computer program 82 may be divided into one or more modules / units, which are stored in the memory 81 and executed by the processor 80 to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program 82 in the vehicle trajectory prediction device 8.

[0111] The vehicle trajectory prediction device may include, but is not limited to, a processor 80 and a memory 81. Those skilled in the art will understand that... Figure 8 This is merely an example of a vehicle trajectory prediction device 8 and does not constitute a limitation on the vehicle trajectory prediction device 8. It may include more or fewer components than shown, or combine certain components, or different components. For example, the vehicle trajectory prediction device may also include input / output devices, network access devices, buses, etc.

[0112] The processor 80 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0113] The memory 81 can be an internal storage unit of the vehicle trajectory prediction device 8, such as a hard disk or memory of the vehicle trajectory prediction device 8. The memory 81 can also be an external storage device of the vehicle trajectory prediction device 8, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the vehicle trajectory prediction device 8. Furthermore, the memory 81 can include both internal and external storage units of the vehicle trajectory prediction device 8. The memory 81 is used to store the computer program and other programs and data required by the vehicle trajectory prediction device. The memory 81 can also be used to temporarily store data that has been output or will be output.

[0114] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0115] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0116] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0117] In the embodiments provided in this application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0118] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0119] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0120] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0121] In addition, this application also provides a computer program product that, when run on a computer, causes the computer to execute the methods in the above-described implementations.

[0122] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A vehicle trajectory prediction method, characterized by, The method comprises: acquiring a lane set related to a planned path of a vehicle; finding relevant vehicles of the vehicle in the lane set, and acquiring vehicle states of the vehicle and the relevant vehicles; deducing a preliminary prediction trajectory of the vehicle within a preset time length according to the vehicle state of the vehicle, and deducing a preliminary prediction trajectory of the relevant vehicle within the preset time length according to the vehicle state of the relevant vehicle; generating a collision detection result by performing collision detection according to the preliminary prediction trajectories of the relevant vehicle and the vehicle within the preset time length; determining a vehicle trajectory prediction result of the vehicle according to the collision detection result.

2. The method of claim 1, wherein, determining a vehicle trajectory prediction result of the vehicle according to the collision detection result; when the collision detection result is that at least one of the vehicle and all the relevant vehicles has a collision risk within the prediction time length, inputting the preliminary prediction trajectories of the relevant vehicles and the vehicle within the preset time length into a preset intelligent driver model to update the vehicle trajectory prediction result of the vehicle; when the collision detection result is that the vehicle has no collision risk with all the relevant vehicles within the preset time length, determining the preliminary prediction trajectory of the vehicle within the preset time length as the vehicle trajectory prediction result.

3. The method of claim 1, wherein, acquiring a lane set related to a planned path of a vehicle, comprising: acquiring a lane in which the vehicle is currently located; and acquiring lanes involved in the planned path of the vehicle.

4. The method of claim 1, wherein, finding relevant vehicles of the vehicle in the lane set, comprising: acquiring a current driving lane of a surrounding vehicle of the vehicle; when the current driving lane and a lane in the lane set have an intersection, determining that the vehicle belongs to the relevant vehicles of the vehicle.

5. The method of claim 1, wherein, finding relevant vehicles of the vehicle in the lane set, comprising: acquiring a motion speed and a motion direction of a surrounding vehicle of the vehicle, predicting an end position after a preset time length, and determining a motion line segment according to a current position of the vehicle and the end position; generating a polygon in which the vehicle is located according to a boundary line of a lane through which the vehicle passes within the preset time length; when the motion line segment intersects with the polygon, determining that the vehicle belongs to the relevant vehicles of the vehicle.

6. The method of claim 1, wherein, deducing a preliminary prediction trajectory of the vehicle within a preset time length according to a vehicle state of the vehicle, comprising: when the vehicle state of the vehicle is a keep intention, deducing a preliminary prediction trajectory of the vehicle within the preset time length as a constant speed motion trajectory or a constant acceleration motion trajectory along a current lane center line; when the vehicle state of the vehicle is a lane change intention, deducing a preliminary prediction trajectory of the vehicle within the preset time length as a transition trajectory from a current lane to a target lane.

7. The method of claim 1, wherein, deducing a preliminary prediction trajectory of the relevant vehicle within a preset time length according to a vehicle state of the relevant vehicle, comprising: when a distance between the position of the relevant vehicle and a center of a lane in which the relevant vehicle is located is less than a first preset distance, and an angle between a speed direction of the relevant vehicle and a lane line of the lane in which the relevant vehicle is located is less than a first preset angle, determining that the relevant vehicle has a keep intention, and deducing a keep trajectory of the relevant vehicle based on the distance by which the relevant vehicle deviates from the center of the lane. in the speed direction of the related vehicle and the lane line is greater than a first preset angle and less than a second preset angle, it is determined that the related vehicle has a lane-changing intention, and a lane-changing trajectory of the related vehicle is deduced according to a center line of a current lane of the related vehicle and a target lane of lane-changing; in the speed direction of the related vehicle and the lane line is greater than the second preset angle and less than a third preset angle, it is determined that the related vehicle has a crossing intention, and a crossing trajectory of the related vehicle is deduced according to a motion trend of the related vehicle; in the speed direction of the related vehicle and the speed direction of the ego vehicle is greater than the third preset angle, it is determined that the related vehicle has a head-on intention, and a head-on trajectory of the related vehicle is deduced according to an opposite direction of a center line of a lane where the related vehicle is located, and a lane line direction of the ego vehicle is the same as a driving direction of the ego vehicle.

8. A vehicle trajectory prediction device characterized by comprising: The device comprises: a lane set acquisition unit configured to acquire a lane set related to a planned path of an ego vehicle; a related vehicle searching unit configured to search for a related vehicle of the ego vehicle in the lane set, and acquire vehicle states of the ego vehicle and the related vehicle; a preliminary predicted trajectory deduction unit configured to deduce a preliminary predicted trajectory of the ego vehicle within a preset time length according to the vehicle state of the ego vehicle, and deduce a preliminary predicted trajectory of the related vehicle within the preset time length according to the vehicle state of the related vehicle; a collision detection unit configured to perform collision detection according to the preliminary predicted trajectories of the related vehicle and the ego vehicle within the preset time length to generate a collision detection result; a vehicle trajectory prediction result determination unit configured to determine a vehicle trajectory prediction result of the ego vehicle according to the collision detection result. 9.A vehicle trajectory prediction device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program, so that the vehicle trajectory prediction device implements the method in any one of claims 1-7.

10. A computer program product comprising computer program instructions, characterised in that, The computer program is run, so that the method in any one of claims 1-7 is executed.

Images