Intelligent vehicle trajectory planning method based on space-time risk map and related device

Abstract

This application discloses a method and related apparatus for intelligent vehicle trajectory planning based on a spatiotemporal risk map, relating to the field of intelligent transportation technology. The method includes generating a spatiotemporal risk sub-map for a given road area based on obstacle vehicle information and the movement trajectory of the obstacle vehicle within that road area. Furthermore, it generates a first risk sub-map and a second risk sub-map for the given road area based on road boundary line distribution information and crossable lane line distribution information, thereby determining the spatiotemporal risk map for the given road area. Then, based on a reference trajectory and the spatiotemporal risk map, a dynamic safety corridor is generated. Finally, based on the dynamic safety corridor, an updated optimal control model is solved to obtain the optimal planned trajectory for the target intelligent vehicle. This application achieves a unified description of different heterogeneous traffic elements from a risk perspective, effectively reducing the complexity of trajectory planning, improving the efficiency of trajectory planning, and reducing the consumption of computational resources.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, and in particular to an intelligent vehicle trajectory planning method and related device based on a spatiotemporal risk map. Background Technology

[0002] Autonomous driving is considered a major innovative technology with the potential to play a positive role in reducing traffic accidents, alleviating traffic congestion, and reducing energy consumption. In autonomous driving systems, the planning layer, as a crucial component connecting the perception and control layers, is considered a direct reflection of the intelligence level of connected vehicles. Its objective is to find a collision-free, comfortable, and smooth trajectory.

[0003] Among various trajectory planning methods for intelligent connected vehicles, optimization-based methods generally describe the trajectory planning problem as the optimal control problem (OCP) in the most accurate, complete, and unified way through constraints and objective functions. They are characterized by high adaptability and flexibility and are widely used in the field of spatiotemporal trajectory planning. However, this process requires significant computational resources in complex urban environments with dense obstacles. This is because complex urban environments contain many heterogeneous traffic elements, such as dynamic obstacles and static obstacles (e.g., lane boundaries), each with different mathematical descriptions, such as constraints and costs. This complexity makes OCP a computationally difficult non-convex optimization problem. Furthermore, high-precision maps and prediction modules provide rich traffic scene information, including reference lanes, occupancy maps, and probabilistic multimodal future trajectories. This information can help intelligent connected vehicles plan driving trajectories, but it also increases the complexity of the planning process, further exacerbating the consumption of computational resources. In summary, current autonomous driving systems struggle to directly understand and efficiently utilize this information. Summary of the Invention

[0004] The purpose of this application is to provide an intelligent vehicle trajectory planning method and related device based on a spatiotemporal risk map, which can realize a unified description of different heterogeneous traffic elements, effectively reduce the complexity of trajectory planning, improve the efficiency of trajectory planning, and reduce the consumption of computing resources.

[0005] To achieve the above objectives, this application provides the following solution:

[0006] Firstly, this application provides an intelligent vehicle trajectory planning method based on a spatiotemporal risk map, including:

[0007] The system acquires information on the distribution of road boundary lines, the distribution of crossable lane lines, and information on obstacle vehicles in a certain road area surrounding the target intelligent vehicle, and predicts the movement trajectory of the obstacle vehicle in the certain road area based on the obstacle vehicle information.

[0008] Based on the information of the obstructing vehicles and their movement trajectories in a certain road area, a spatiotemporal risk sub-map of the certain road area is generated.

[0009] Based on the road boundary line distribution information and the crossable lane line distribution information, a first risk sub-map and a second risk sub-map for the certain road area are generated respectively.

[0010] Based on the spatiotemporal risk sub-map, the first risk sub-map, and the second risk sub-map, a spatiotemporal risk map for the certain road area is determined.

[0011] Using a dynamic programming algorithm, a reference trajectory of the target intelligent vehicle in a certain road area is obtained, and a dynamic safety corridor is generated based on the reference trajectory and the spatiotemporal risk map.

[0012] Based on the dynamic safety corridor, the boundary constraints of the dynamic safety corridor in the optimal control model are updated, and the updated optimal control model is solved to obtain the optimal planned trajectory of the target intelligent vehicle.

[0013] In a second aspect, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent vehicle trajectory planning method based on a spatiotemporal risk map as described in the first aspect.

[0014] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent vehicle trajectory planning method based on a spatiotemporal risk map as described in the first aspect.

[0015] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the intelligent vehicle trajectory planning method based on a spatiotemporal risk map as described in the first aspect.

[0016] According to the specific embodiments provided in this application, this application has the following technical effects:

[0017] This application provides a method and related apparatus for intelligent vehicle trajectory planning based on a spatiotemporal risk map. Based on obstacle vehicle information and the movement trajectory of the obstacle vehicle within a certain road area, a spatiotemporal risk sub-map of the certain road area is generated. Furthermore, based on road boundary line distribution information and traversable lane line distribution information, a first risk sub-map and a second risk sub-map of the certain road area are generated. Then, based on the spatiotemporal risk sub-map, the first risk sub-map, and the second risk sub-map, the spatiotemporal risk map of the certain road area is determined. Next, based on the reference trajectory and the spatiotemporal risk map, a dynamic safety corridor is generated. Based on the dynamic safety corridor, an updated optimal control model is solved to obtain the optimal planned trajectory of the target intelligent vehicle. This application, through the above scheme, unifies dynamic and static heterogeneous traffic elements such as obstacle vehicles, road boundary lines, and traversable lane lines into the risk of a road area, and ultimately generates a spatiotemporal risk map corresponding to the road area. This achieves a unified description of different heterogeneous traffic elements, effectively reduces the complexity of trajectory planning, improves the efficiency of trajectory planning, reduces the consumption of computing resources, and effectively ensures the real-time requirements of trajectory planning. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments 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.

[0019] Figure 1 This is a flowchart illustrating an intelligent vehicle trajectory planning method based on a spatiotemporal risk map according to an embodiment of this application.

[0020] Figure 2 This is a flowchart illustrating another intelligent vehicle trajectory planning method based on a spatiotemporal risk map in one embodiment of this application;

[0021] Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] In one exemplary embodiment, such as Figure 1 As shown, an intelligent vehicle trajectory planning method based on a spatiotemporal risk map is provided, including the following steps 101 to 106. Wherein:

[0025] Step 101: Obtain information on the distribution of road boundary lines, the distribution of crossable lane lines, and information on obstacle vehicles in a certain road area around the target intelligent vehicle, and predict the movement trajectory of the obstacle vehicle in the certain road area based on the obstacle vehicle information.

[0026] Step 102: Generate a spatiotemporal risk sub-map of the certain road area based on the obstacle vehicle information and the movement trajectory of the obstacle vehicle in the certain road area.

[0027] Step 103: Generate a first risk sub-map and a second risk sub-map for the certain road area based on the road boundary line distribution information and the crossable lane line distribution information, respectively.

[0028] Step 104: Determine the spatiotemporal risk map of the certain road area based on the spatiotemporal risk sub-map, the first risk sub-map, and the second risk sub-map.

[0029] Step 105: Using a dynamic programming algorithm, a reference trajectory of the target intelligent vehicle in a certain road area is obtained, and a dynamic safety corridor is generated based on the reference trajectory and the spatiotemporal risk map.

[0030] Step 106: Based on the dynamic safety corridor, update the dynamic safety corridor boundary constraints in the optimal control model, and solve the updated optimal control model to obtain the optimal planned trajectory of the target intelligent vehicle.

[0031] As an optional implementation, step 105 generates a dynamic safety corridor based on the reference trajectory and the spatiotemporal risk map, specifically including:

[0032] Step 105.11: Determine a preset number of initial rectangular regions based on all adjacent discrete trajectory points on the reference trajectory; the preset number is equal to the number of pairs of all adjacent discrete trajectory points; each pair of adjacent discrete trajectory points is a diagonal point of a corresponding initial rectangular region.

[0033] Step 105.12: Perform several rounds of expansion on each initial rectangular region to obtain the expanded rectangular region corresponding to each initial rectangular region.

[0034] Step 105.13: Determine a dynamic safety corridor based on a preset number of the expanded rectangular regions; the dynamic safety corridor is composed of a preset number of the expanded rectangular regions.

[0035] In step 105.12, the expansion process of each initial rectangular region in several rounds is as follows:

[0036] According to the preset translation distance increment, the four sides of the initial rectangular region are translated outward once in sequence to obtain the first round of expanded rectangular region. Then, according to the preset translation distance increment, the four sides of the first round of expanded rectangular region are translated outward once in sequence to obtain the second round of expanded rectangular region. The expansion operation is repeated until all sides of the expanded rectangular region are marked as expanded. The translation direction of each side is perpendicular to each side.

[0037] The process of performing an outward translation operation on each edge is as follows:

[0038] 1) Determine whether each edge is an edge that has been fully expanded, and obtain the first determination result.

[0039] 2) When the first judgment result is yes, then abandon the outward translation operation for each edge.

[0040] 3) When the first judgment result is negative, each side is translated outward once according to the preset translation distance increment to obtain the corresponding rectangular area.

[0041] 4) Determine whether the corresponding rectangular area contains a location point with a risk value higher than a preset risk threshold, and obtain a second determination result.

[0042] 5) When the second judgment result is yes, abandon the translation operation performed on each edge and mark each edge as an edge that has been expanded.

[0043] 6) If the second judgment result is negative, retain the translation operation performed on each edge.

[0044] As an optional implementation, step 102 specifically includes:

[0045] Based on the information of the obstructing vehicles and their movement trajectories in a certain road area, a spatiotemporal risk sub-map of the certain road area is generated using an obstructing vehicle potential field model; the mathematical expression of the obstructing vehicle potential field model (the potential field model of obstructing vehicle i) is as follows:

[0046]

[0047] Wherein, vector EV This represents the risk posed by the obstacle vehicle to the surrounding environment; δ represents any point j(x) in space. j ,y j The angle formed clockwise between the direction of motion of the obstacle vehicle and the direction of motion of the obstacle vehicle; a is the current acceleration of the obstacle vehicle; c is the current acceleration of the obstacle vehicle. v M is a constant that limits the maximum risk to a vehicle. i The virtual mass of the target intelligent vehicle (affected by vehicle type and actual mass) The distance is the elliptical distance.

[0048] The elliptic distance The mathematical expression is as follows:

[0049]

[0050] Where, τ x and τ y The parameters are ellipse parameters, representing the safety margins of the obstacle vehicle in the driving direction and the perpendicular direction (perpendicular to the driving direction), respectively. They are related to the size and speed of the obstacle vehicle, and the specific calculation formulas are as follows:

[0051]

[0052] Among them, L and L b These represent the length and width of the obstacle vehicle, respectively; ζ x and ζ y The adjustment parameter for safety margin is determined by safety requirements in different directions; λ is a speed-related weighting coefficient that can adjust the impact range of the risk posed by the (obstacle) vehicle in the direction of travel; v i The speed of the obstacle vehicle; The coordinates of the center of mass of the obstacle vehicle are given.

[0053] Because the safety requirements for obstacle vehicles differ depending on whether they are traveling in the same direction as the vehicle or perpendicular to it—for example, it is generally more acceptable to be relatively close to the obstacle vehicle in the direction perpendicular to its travel direction than to be close to it in its own direction—this embodiment introduces elliptical distance instead of actual distance to model the field strength.

[0054] As an optional implementation, step 103, based on the road boundary line distribution information, generates a first risk sub-map for the specified road area, specifically including:

[0055] Based on the road boundary line distribution information, a first risk sub-map of the given road area is generated using the road boundary line potential field model; the mathematical expression of the road boundary line potential field model (potential field model of road boundary line i) is as follows:

[0056]

[0057] Wherein, vector E B Indicates the risks arising from road boundary lines; Denotes the distance from the boundary line of the i-th road to any point j(x) in space. j ,y j The distance vector of ); c b A is a constant that limits the maximum risk at the road boundary; B This refers to the field strength control parameter, i.e., the parameter that controls the magnitude of the field strength; D max This is the affected distance threshold, i.e., the distance threshold of the road boundary line's influence.

[0058] As an optional implementation, step 103 generates a second risk sub-map for the specified road area based on the traversable lane distribution information, specifically including:

[0059] Based on the distribution information of the crossable lane lines, a second risk sub-map for the given road area is generated using the crossable lane line potential field model. The actual impact of the crossable lane markings is defined according to traffic regulations. The mathematical expression of the crossable lane line potential field model (potential field model of crossable lane line i) is as follows:

[0060]

[0061] Wherein, vector E L Indicates the risk posed by crossing lane lines; A L The parameter is used to adjust the peak field strength; σ is a constant that determines the rate of increase of field strength when approaching the lane line. Denotes the i-th crossable lane line leading to any point j(x) in space. j ,y j The distance vector of ).

[0062] As an optional implementation, the updated optimal control model includes an objective function, updated dynamic safety corridor boundary constraints, vehicle system dynamics constraints, and vehicle dynamics boundary constraints; solving the updated optimal control model to obtain the optimal planned trajectory of the target intelligent vehicle specifically includes:

[0063] Based on the reference trajectory, the updated optimal control model is solved with the objective function as the goal, to obtain the optimal planned trajectory of the target intelligent vehicle.

[0064] As an optional implementation, the mathematical expression of the objective function is as follows:

[0065]

[0066] in, Denotes the objective function; φ i (t) represents the front wheel steering angle of the target intelligent vehicle at time t, and β i (t) represents the angular velocity of the front wheel steering angle of the target intelligent vehicle at time t. i (t) represents the rate of change of the acceleration of the target intelligent vehicle at time t, L W Let be the wheelbase of the target intelligent vehicle. These are the weighting coefficients for each sub-indicator; Smoothness expression: For trajectory smoothness; Comfort expression: For passenger comfort; t0 is the initial time of the reference trajectory, t T The end time of the reference trajectory.

[0067] To help understand the above-mentioned solution of this application, the technical solution of this application will be further explained below from the perspective of solution design.

[0068] This application addresses the shortcomings of existing technologies by proposing an intelligent vehicle trajectory planning method based on a spatiotemporal risk map. It models the risks of traffic elements in different states or types based on potential field theory and establishes a spatiotemporal risk map based on the risk mapping. Then, combined with a dynamic safety corridor, it solves an optimal control problem. This method supports real-time trajectory planning for intelligent connected vehicles in highly dynamic and high-risk scenarios.

[0069] like Figure 2 As shown, the design concept of the intelligent vehicle trajectory planning method based on spatiotemporal risk maps in this application is as follows:

[0070] Step 1: Perform risk modeling on traffic elements in different states, and establish a spatiotemporal risk map based on the risk mapping.

[0071] (1.1) A potential field model is established for the road boundary line, which is defined as being generated by the road boundary line and not changing with time. The constructed potential field model for the road boundary line is described above.

[0072] (1.2) A potential field model is established for crossable lane markings. These lane markings guide intelligent connected vehicles (such as target intelligent vehicles) to travel along the lane centerline and create slight obstacles during lane changes. The actual impact of crossable lane markings is defined according to traffic regulations. The potential field model for crossable lane markings is described above.

[0073] (1.3) Because the safety requirements of the obstacle vehicle differ in the direction of vehicle travel and perpendicular to the direction of vehicle travel, for example, it is generally more acceptable to be relatively close to the obstacle vehicle in the direction perpendicular to the direction of vehicle travel than to be close to it in its own direction of travel. Therefore, elliptical distance is introduced to model the field strength instead of the actual distance. The specific calculation formula for elliptical distance is given above.

[0074] (1.4) Establish a potential field model for the obstacle vehicle, taking into account its motion state and spatial position. For the specific potential field model of the obstacle vehicle, please refer to the previous text.

[0075] (1.5) Based on the road boundaries (lines), crossable lane lines, and obstacle vehicle potential fields established in steps 1.1-1.4, calculate the risk value of each location point in the spatiotemporal risk map. Considering the independence of each part, local normalization can be performed on each part of the driving environment. Thus, the risk value of each location in the spatiotemporal risk map is recorded as:

[0076]

[0077] Among them, |E T | * It is the magnitude of the total field strength at any point in space; |E B_max |is|E B The maximum value in |E; L_max |is|E L The maximum value in |E; V_max |is|E V The maximum value in |; ω b ω l ω v This is the proportionality coefficient.

[0078] (1.6) Combine location (x, y), time t, and risk |E T | * Matching and mapping are performed to form a spatiotemporal risk map, defined as:

[0079]

[0080] Spatiotemporal risk maps can quantify the risk level at each moment and location.

[0081] Step 2: Based on the spatiotemporal risk map generated in Step 1, construct a dynamic safety corridor according to preset risk thresholds.

[0082] (2.1) First, dynamic programming is used to determine the reference trajectory (traj). ref x(t), traj ref y(t), traj ref θ(t)), t∈[t0, t Tx, y, and θ represent the abscissa, ordinate, and heading angle of the reference trajectory in a Cartesian coordinate system. t0 is the initial time of the trajectory to be planned, which is also the initial time of the reference trajectory; t T This is the end time of the trajectory to be planned, and also the end time of the reference trajectory. Reference trajectory traj ref Discretized into N ref N trajectory points (also called path points) ref The set of trajectory points is represented as Trajectory point time interval [t] i , t i+1 ],

[0083] (2.2) Connect two consecutive path points q i With q i+1 Consider the two diagonal points of the initial rectangle (i.e., the initial rectangular region) to generate a rectangular initial channel. The initial rectangle may fail to be created in certain cases, for example, q. i With q i+1 The line connecting q is parallel to the X-axis (i.e., the x-coordinate in a Cartesian coordinate system). Only when q i With q i+1 For the initial rectangle to be successfully created, the connecting lines between the points must not be parallel to the X-axis. Furthermore, to ensure the success of the final dynamic safety corridor construction process, the risk value within the initial rectangle constructed from continuous trajectory points must be below a preset risk threshold. If either of these conditions is not met, additional steps must be performed to find the initial rectangle. For example, if the connecting lines between two adjacent path points are parallel to the X-axis, either path point can be moved a certain distance along the Y-axis (i.e., the ordinate in a Cartesian coordinate system) so that the connecting lines are no longer parallel to the X-axis. The same method can be used to handle another scenario.

[0084] (2.3) Gradually expand the initial rectangle In a Cartesian coordinate system, starting from an initial rectangle, the coordinates are progressively expanded in the order of +y, +x, -y, -x with a step size Δd (i.e., a preset translation distance increment). A preset risk threshold R is used. thr This is a condition for the expansion to complete. Specifically, if the risk in a certain direction is higher than R in the Kth step. thr If a certain condition is met, expansion in that direction is prohibited, and starting from step K+1, expansion will only occur in the remaining directions in the specified order. Expansion stops when all directions are prohibited.

[0085] (2.4) Extensions that are not disabled are merged into the local security corridor. (i.e., the expanded rectangular region i) It is a rectangle. These are the upper and lower boundaries along the x-axis and y-axis, respectively.

[0086] (2.5) All local security corridors Along the reference trajectory traj ref They combine to form a dynamic safety corridor.

[0087] Step 3: Considering trajectory smoothness and passenger comfort, construct an optimal control problem and solve it to generate a corridor trajectory that meets the requirements.

[0088] (3.1) The objective function considers two aspects: trajectory smoothness and passenger comfort. Trajectory smoothness refers to the smoothness of trajectory changes during motion and is described as curvature-related. Passenger comfort settings are related to control variables, encouraging them to approach 0, which means reducing drastic changes or vibrations. The definition of the objective function is given above.

[0089] (3.2) The target intelligent vehicle travels within the dynamic safety corridor, ensuring its safe avoidance of all dynamic and static obstacles. Considering the geometry of the target intelligent vehicle, it is necessary to ensure that the entire rectangular body of the target intelligent vehicle lies within the boundary of the dynamic safety corridor. A double-circle model is used to describe the shape of the target intelligent vehicle, with two circles uniformly covering the vehicle body. The centers of the two circles... and It is the quarter point along the axle. The center position is determined using the following formula:

[0090]

[0091]

[0092] Where, θ i (t) represents the heading angle of the target intelligent vehicle at time t; α r The coefficients are defined for the geometric parameters of the target intelligent vehicle, where r is the radius of the circle, and they can be obtained through the size parameter L. w L r L f To calculate. Additionally, L w L is the wheelbase. f For front overhang distance, L r Rear overhang distance, L b x is the width of the vehicle; i (t) and y i (t) represents the centroid coordinate of the target intelligent vehicle at time t, where the centroid of the target intelligent vehicle is the center of the line connecting the centers of its two rear wheels.

[0093] (3.3) Establish dynamic safety corridor boundary constraints, which serve as collision constraints in the optimal control problem, and are expressed as follows:

[0094]

[0095] The aforementioned collision constraints ensure that the target intelligent vehicle at time t will not exceed the local safety corridor. The boundary.

[0096] (3.4) During trajectory planning, each variable needs to follow (not violate) the evolutionary relationship of the vehicle's kinematics. Therefore, vehicle system dynamic constraints are established, expressed as:

[0097]

[0098] Among them, (x i (t), y i (t) represents the position coordinates of the target intelligent vehicle; v i (t) represents the vehicle speed along the longitudinal axis of the vehicle body; θ i (t), φ i (t) represent the heading angle and front wheel steering angle in the Cartesian coordinate system, respectively; a i (t) represents the acceleration along the longitudinal axis of the vehicle body; jerk i (t) represents the rate of change of acceleration; β i (t) represents the angular velocity of the front wheel. Furthermore, χ i (t) is the state variable of the target intelligent vehicle at time t; U i (t) is the control variable of the target intelligent vehicle at time t.

[0099] (3.5) The above χ i (t) and U i (t) has upper and lower bounds, reflecting physical or mechanical limitations related to vehicle kinematics, thus establishing vehicle dynamics boundary constraints, expressed as:

[0100]

[0101] Where, φ i,min φ i,max φ i The upper and lower boundaries of (t); β i,min β i,max Indicates β i The upper and lower boundaries of (t); v i,min v i,max Indicates v i The upper and lower boundaries of (t); a i,min a i,max Indicates a i The upper and lower boundaries of (t); jerk i,min jerk i,max jerki The upper and lower boundaries of (t).

[0102] (3.6) Taking into account the objective function, collision constraints, vehicle system dynamics constraints, and vehicle dynamics boundary constraints, an optimal control model is constructed, expressed as:

[0103]

[0104]

[0105] The optimal planning trajectory is solved using the above optimal control model.

[0106] (3.7) Using the reference trajectory as the initial iteration value, solve the optimal control model to output the optimal planned trajectory.

[0107] This application also provides an application scenario in which the above-mentioned intelligent vehicle trajectory planning method based on a spatiotemporal risk map is applied. Specifically, the intelligent vehicle trajectory planning method based on a spatiotemporal risk map provided in this embodiment can be applied in an autonomous driving scenario. An autonomous driving scenario includes a surrounding road condition information acquisition stage, a trajectory planning stage, and an autonomous driving stage. Road boundary line distribution information, traversable lane line distribution information, and obstacle vehicle information enter the trajectory planning stage from the surrounding road condition information acquisition stage. In the trajectory planning stage, an optimal planned trajectory is generated based on the road boundary line distribution information, traversable lane line distribution information, and obstacle vehicle information. The autonomous driving stage then drives the target intelligent vehicle based on the optimal planned trajectory. The intelligent vehicle trajectory planning method based on a spatiotemporal risk map provided in this embodiment belongs to the trajectory planning stage. Specifically, in the trajectory planning stage, trajectory planning for a future period is performed based on the road boundary line distribution information, traversable lane line distribution information, and obstacle vehicle information.

[0108] This application provides a novel trajectory planning method for intelligent connected vehicles, which can uniformly describe different traffic elements in complex environments, thereby integrating traffic scene information and reducing the impact of heterogeneous element types, combinations, and the motion state of obstacle vehicles on the real-time performance and quality of planned trajectory generation.

[0109] Specifically, this application has the following advantages:

[0110] (1) The intelligent vehicle trajectory planning method based on spatiotemporal risk map proposed in this application relies on risk space in the planning process. By quantifying and integrating risk factors in the environment, it can more flexibly respond to dynamic changes in complex traffic situations and realize the transformation from "spatial safety" to "risk safety".

[0111] (2) The spatiotemporal risk map generation method based on potential field proposed in this application can uniformly describe complex traffic environments (heterogeneous traffic elements), map environmental information to risk space, and enable complex environmental information to be uniformly described and evaluated.

[0112] (3) This application proposes a method for constructing dynamic safety corridors. Dynamic safety corridors are generated based on spatiotemporal risk maps. Unlike corridors that guarantee a free space within the environment, the dynamic safety corridors constructed by the method in this application can ensure that the risk within the corridor remains below a preset risk threshold for a period of time. Furthermore, by adjusting the preset risk threshold, personalized driving needs can be effectively addressed.

[0113] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 3 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores trajectory planning-related data. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements an intelligent vehicle trajectory planning method based on a spatiotemporal risk map.

[0114] Those skilled in the art will understand that Figure 3 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0115] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0116] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0117] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0118] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0119] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0120] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0121] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for intelligent vehicle trajectory planning based on a spatiotemporal risk map, characterized in that, The intelligent vehicle trajectory planning method based on spatiotemporal risk maps includes: The system acquires information on the distribution of road boundary lines, the distribution of crossable lane lines, and information on obstacle vehicles in a certain road area around the target intelligent vehicle, and predicts the movement trajectory of the obstacle vehicle in the certain road area based on the obstacle vehicle information. Based on the information of the obstructing vehicles and the movement trajectory of the obstructing vehicles in the certain road area, a spatiotemporal risk sub-map of the certain road area is generated. Based on the road boundary line distribution information and the crossable lane line distribution information, a first risk sub-map and a second risk sub-map for the certain road area are generated respectively. Based on the spatiotemporal risk sub-map, the first risk sub-map, and the second risk sub-map, determine the spatiotemporal risk map of the certain road area; Using a dynamic programming algorithm, a reference trajectory of the target intelligent vehicle in a certain road area is obtained, and a dynamic safety corridor is generated based on the reference trajectory and the spatiotemporal risk map. Based on the dynamic safety corridor, update the dynamic safety corridor boundary constraints in the optimal control model, and solve the updated optimal control model to obtain the optimal planned trajectory of the target intelligent vehicle. Specifically, based on the obstacle vehicle information and the movement trajectory of the obstacle vehicle in the defined road area, a spatiotemporal risk sub-map of the defined road area is generated, including: Based on the information of the obstructing vehicles and their movement trajectories in a specific road area, a spatiotemporal risk sub-map of the specific road area is generated using an obstructing vehicle potential field model. The mathematical expression of the obstructing vehicle potential field model is as follows: ； Where, vector This indicates the risk that the obstructing vehicle poses to the surrounding environment; any point in space The clockwise angle formed with the direction of movement of the obstacle vehicle; The current acceleration of the vehicle with the obstacle; It is a constant. The virtual mass of the target intelligent vehicle. The distance is the elliptical distance. The elliptic distance The mathematical expression is as follows: ； in, and Let be the ellipse parameters, representing the safety margins of the obstacle vehicle in the driving direction and the vertical direction, respectively. The calculation formula is as follows: ； ； in, and These represent the length and width of the obstacle vehicle, respectively; and To adjust the parameters; These are the weighting coefficients; The speed of the obstacle vehicle; , Let be the coordinates of the centroid of the obstacle vehicle; Based on the road boundary line distribution information, a first risk sub-map of the specified road area is generated, specifically including: Based on the road boundary line distribution information, a first risk sub-map of the given road area is generated using a road boundary line potential field model; the mathematical expression of the road boundary line potential field model is as follows: ； Where, vector Indicates the risks arising from road boundary lines; Indicates the first The road boundary line to any point in space The distance vector; It is a constant; These are the field strength control parameters; Distance threshold; Based on the traversable lane distribution information, a second risk sub-map for the specified road area is generated, specifically including: Based on the crossable lane line distribution information, a second risk sub-map for the given road area is generated using the crossable lane line potential field model; the mathematical expression of the crossable lane line potential field model is as follows: ； Where, vector This indicates the risk of crossing lane lines; To adjust the parameters for the peak electric field strength; It is a constant; Indicates the first A lane line that can cross to any point in space The distance vector.

2. The intelligent vehicle trajectory planning method based on spatiotemporal risk map according to claim 1, characterized in that, Based on the reference trajectory and the spatiotemporal risk map, a dynamic safety corridor is generated, specifically including: Based on all adjacent discrete trajectory points on the reference trajectory, a preset number of initial rectangular regions are determined; the preset number is equal to the number of pairs of all adjacent discrete trajectory points; each pair of adjacent discrete trajectory points is a diagonal point of a corresponding initial rectangular region. Each initial rectangular region is expanded in several rounds to obtain the expanded rectangular region corresponding to each initial rectangular region; A dynamic safety corridor is determined based on a preset number of the expanded rectangular regions; the dynamic safety corridor is composed of a preset number of the expanded rectangular regions. The expansion process of each initial rectangular region in several rounds is as follows: According to the preset translation distance increment, the four sides of the initial rectangular region are translated outward once in sequence to obtain the first round of expanded rectangular region. Then, according to the preset translation distance increment, the four sides of the first round of expanded rectangular region are translated outward once in sequence to obtain the second round of expanded rectangular region. The expansion operation is repeated until all sides of the expanded rectangular region are marked as expanded. The translation direction of each side is perpendicular to each side. The process of performing an outward translation operation on each edge is as follows: Determine whether each edge is a fully expanded edge to obtain the first determination result; If the first judgment result is yes, then the outward translation operation for each edge is abandoned; If the first judgment result is negative, then each side is translated outward once according to the preset translation distance increment to obtain the corresponding rectangular area; Determine whether the corresponding rectangular area contains a location point with a risk value higher than a preset risk threshold, and obtain a second determination result; When the second judgment result is yes, abandon the one outward translation operation performed on each edge, and mark each edge as an edge that has been expanded. If the second judgment result is negative, the one outward translation operation performed on each edge is retained.

3. The intelligent vehicle trajectory planning method based on a spatiotemporal risk map according to claim 1, characterized in that, The updated optimal control model includes an objective function, updated dynamic safety corridor boundary constraints, vehicle system dynamics constraints, and vehicle dynamics boundary constraints; solving the updated optimal control model to obtain the optimal planned trajectory of the target intelligent vehicle specifically includes: Based on the reference trajectory, the updated optimal control model is solved with the objective function as the goal, to obtain the optimal planned trajectory of the target intelligent vehicle.

4. The intelligent vehicle trajectory planning method based on a spatiotemporal risk map according to claim 3, characterized in that, The mathematical expression of the objective function is as follows: ； in, Describe the objective function; for The front wheel steering angle of the target intelligent vehicle at that moment. for The angular velocity of the front wheel of the target intelligent vehicle at the specified time. for The rate of change of the acceleration of the target intelligent vehicle at time t. Let be the wheelbase of the target intelligent vehicle. , These are the weighting coefficients for each item; Expression: For smooth trajectory; Expression: For passenger comfort; The initial time of the reference trajectory, The end time of the reference trajectory.

5. A computer 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 to implement the intelligent vehicle trajectory planning method based on a spatiotemporal risk map as described in any one of claims 1-4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the intelligent vehicle trajectory planning method based on a spatiotemporal risk map as described in any one of claims 1-4.

7. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the intelligent vehicle trajectory planning method based on a spatiotemporal risk map as described in any one of claims 1-4.

Images