Urban scene automatic driving compliance spatiotemporal motion planning method and device

Abstract

This application relates to the field of path planning technology, and in particular to a method and apparatus for planning the spatiotemporal motion of autonomous driving in urban scenarios to ensure compliance. The method includes: acquiring urban road traffic regulations and highway traffic regulations; parsing the urban road traffic regulations and highway traffic regulations to generate compliance constraint information in urban scenarios; constructing spatiotemporal semantic corridor information of autonomous vehicles based on the compliance constraint information in urban scenarios; constructing a compliance spatiotemporal feasible domain of autonomous vehicles based on the spatiotemporal semantic corridor information; constructing a safety potential field of autonomous vehicles based on the compliance spatiotemporal feasible domain; and planning the compliance target motion trajectory of autonomous vehicles based on the gradient of the safety potential field and the dynamic model of autonomous vehicles. This achieves the real-time generation of autonomous driving trajectories that strictly comply with traffic regulations while being dynamically feasible, safe, and smooth in complex urban scenarios.

Description

Technical Field

[0001] This application relates to the field of path planning technology, and in particular to a method and apparatus for compliant spatiotemporal motion planning for autonomous driving in urban scenarios. Background Technology

[0002] Compliance-compliant spatiotemporal motion planning technology plays a crucial role in the overall technological framework of autonomous driving. Currently, mainstream autonomous driving motion planning methods are based on search, sampling, optimization, and learning. Regarding compliance, the planned paths for autonomous driving must strictly adhere to traffic regulations, including but not limited to speed limits, lane rules, and right-of-way. In terms of generating feasible regions, autonomous driving planning needs to consider the vehicle's kinematic characteristics to generate dynamic feasible regions, ensuring the smoothness and safety of the planned path. Therefore, there is an urgent need for a motion planning method that can simultaneously meet the requirements of compliance, spatiotemporal semantic corridor generation, and real-time performance in complex urban environments. Summary of the Invention

[0003] This application provides a spatiotemporal motion planning method and device for autonomous driving compliance in urban scenarios, which realizes the real-time generation of autonomous driving trajectories that strictly comply with traffic regulations, are dynamically feasible, safe and smooth in complex urban scenarios.

[0004] The first aspect of this application provides a method for spatiotemporal motion planning for autonomous driving compliance in urban scenarios, comprising the following steps: acquiring urban road traffic regulations text and highway traffic regulations text; parsing the urban road traffic regulations text and highway traffic regulations text to generate compliance constraint information in urban scenarios; constructing spatiotemporal semantic corridor information of autonomous vehicles based on the compliance constraint information in urban scenarios; constructing a compliance spatiotemporal feasible domain of autonomous vehicles based on the spatiotemporal semantic corridor information; constructing a safety potential field of autonomous vehicles based on the compliance spatiotemporal feasible domain; and planning the compliance target motion trajectory of autonomous vehicles based on the gradient of the safety potential field and the dynamic model of autonomous vehicles.

[0005] Optionally, compliance constraint information in urban scenarios includes one or more combinations of speed limit regulations, safe distance constraints between vehicles, lane changing behavior regulations, right-of-way determination criteria, and stop-and-yield constraints for autonomous vehicles.

[0006] Optionally, the spatiotemporal semantic corridor information includes obstacle-type semantic elements and constraint-type semantic elements. The obstacle-type semantic elements include at least one of static and dynamic obstacle constraints and dynamic obstacle constraints. The constraint-type semantic elements include at least one of traffic light constraints, spatiotemporal semantic corridor driving distance constraints, and vehicle bounding box vertex constraints.

[0007] Optionally, the objective function for the compliance spatiotemporal feasible region is:

[0008]

[0009] Where t represents time, and κ1 and κ2 represent weight coefficients. This represents the cost of the corridor size at time t. This represents the cost of the corridor center point at time t.

[0010] Optionally, the first derivative h of the safety potential field s and the second derivative h ss for:

[0011]

[0012] Where, ρ safe D represents the safety potential field value. safe denoted by , a1 represents the normalized distance of the safety potential field, b1 represents the control potential field strength, and b1 represents the control decay rate.

[0013] Optionally, the gradient of the safety potential field can be calculated using the following formula:

[0014]

[0015] in, This represents the longitudinal component of the repulsive force (X-axis direction). S represents the partial derivative of the normalized distance with respect to the vehicle's global coordinate X, where S represents the normalized distance and X represents the X-axis coordinate. This represents the transverse (Y-axis) component of the repulsive force. This represents the partial derivative of the normalized distance with respect to the vehicle's global coordinates Y, where Y represents the Y-axis coordinate. This represents the curvature in the X direction (a measure of the degree of curvature of the potential field in pure longitudinal (X) motion. A larger value indicates a deeper or steeper longitudinal "potential well"). This represents the square of the unit direction vector in the X direction. This represents the second-order partial derivative of the normalized distance with respect to the X-coordinate. This represents the curvature in the Y direction (a measure of the degree of curvature of the potential field in pure transverse (Y) motion; the larger the value, the steeper the transverse "potential wall"). This represents the square of the unit direction vector in the Y direction. Let denote the second partial derivative of the normalized distance with respect to the Y-coordinate. This indicates the cross curvature (a measure of the coupling relationship between longitudinal and lateral motion). This represents the product of the partial derivative of the normalized distance S with respect to the X coordinate and the partial derivative of the normalized distance S with respect to the Y coordinate. This means that the normalized distance S is first derived with respect to Y, and then with respect to X (this is the mixed partial derivative, which measures the direction vector in the X direction when moving along the Y direction). How fast things are changing.

[0016] Optionally, the formula for planning the trajectory of the compliance target is:

[0017]

[0018] x(0)=x 初始 u(0)=u 初始

[0019] x(T)=x 终止 u(T)=u 终止

[0020] in, Let J(x(t), u(t)) represent the objective function for minimizing the generated trajectory, where x(t) represents the state variable, t represents time, u(t) represents the control variable, and J(x(t), u(t)) represents the cost function. Let f(x(t),u(t)) represent the time derivative of the state variable x(t), and let f(x(t),u(t)) represent the state equation, where x(0) = x 初始 Denotes the initial state constraint, u(0) = u 初始 Denotes the initial control constraint, x(T) = x 终止 Representing the terminal state constraint, u(T) = u 终止 Terminal control constraints.

[0021] A second aspect of this application provides a spatiotemporal motion planning device for autonomous driving compliance in urban scenarios, comprising: an acquisition module for acquiring urban road traffic regulations text and highway traffic regulations text, parsing the urban road traffic regulations text and highway traffic regulations text to generate compliance constraint information in urban scenarios; a construction module for constructing spatiotemporal semantic corridor information of autonomous vehicles based on the compliance constraint information in urban scenarios, and constructing a compliance spatiotemporal feasible domain of autonomous vehicles based on the spatiotemporal semantic corridor information; and a planning module for constructing a safety potential field of autonomous vehicles based on the compliance spatiotemporal feasible domain, and planning the compliance target motion trajectory of autonomous vehicles based on the gradient of the safety potential field and the dynamic model of autonomous vehicles.

[0022] Optionally, compliance constraint information in urban scenarios includes one or more combinations of speed limit regulations, safe distance constraints between vehicles, lane changing behavior regulations, right-of-way determination criteria, and stop-and-yield constraints for autonomous vehicles.

[0023] Optionally, the spatiotemporal semantic corridor information includes obstacle-type semantic elements and constraint-type semantic elements. The obstacle-type semantic elements include at least one of static and dynamic obstacle constraints and dynamic obstacle constraints. The constraint-type semantic elements include at least one of traffic light constraints, spatiotemporal semantic corridor driving distance constraints, and vehicle bounding box vertex constraints.

[0024] Optionally, the objective function for the compliance spatiotemporal feasible region is:

[0025]

[0026] Where t represents time, and κ1 and κ2 represent weight coefficients. This represents the cost of the corridor size at time t. This represents the cost of the corridor center point at time t.

[0027] Optionally, the first derivative h of the safety potential field s and the second derivative h ss for:

[0028]

[0029] Where, ρ safe D represents the safety potential field value. safe denoted by , a1 represents the normalized distance of the safety potential field, b1 represents the control potential field strength, and b1 represents the control decay rate.

[0030] Optionally, the gradient of the safety potential field can be calculated using the following formula:

[0031]

[0032] in, This represents the longitudinal component of the repulsive force (X-axis direction). S represents the partial derivative of the normalized distance with respect to the vehicle's global coordinate X, where S represents the normalized distance and X represents the X-axis coordinate. This represents the transverse (Y-axis) component of the repulsive force. This represents the partial derivative of the normalized distance with respect to the vehicle's global coordinates Y, where Y represents the Y-axis coordinate. This represents the curvature in the X direction (a measure of the degree of curvature of the potential field in pure longitudinal (X) motion. A larger value indicates a deeper or steeper longitudinal "potential well"). This represents the square of the unit direction vector in the X direction. This represents the second-order partial derivative of the normalized distance with respect to the X-coordinate. This represents the curvature in the Y direction (a measure of the degree of curvature of the potential field in pure transverse (Y) motion; the larger the value, the steeper the transverse "potential wall"). This represents the square of the unit direction vector in the Y direction. Let denote the second partial derivative of the normalized distance with respect to the Y-coordinate. This indicates the cross curvature (a measure of the coupling relationship between longitudinal and lateral motion). This represents the product of the partial derivative of the normalized distance S with respect to the X coordinate and the partial derivative of the normalized distance S with respect to the Y coordinate. This means that the normalized distance S is first derived with respect to Y, and then with respect to X (this is the mixed partial derivative, which measures the direction vector in the X direction when moving along the Y direction). How fast things are changing.

[0033] Optionally, the formula for planning the trajectory of the compliance target is:

[0034]

[0035] x(0)=x 初始 u(0)=u 初始

[0036] x(T)=x 终止 u(T)=u 终止

[0037] in, Let J(x(t), u(t)) represent the objective function for minimizing the generated trajectory, where x(t) represents the state variable, t represents time, u(t) represents the control variable, and J(x(t), u(t)) represents the cost function. Let f(x(t),u(t)) represent the time derivative of the state variable x(t), and let f(x(t),u(t)) represent the state equation, where x(0) = x 初始 Denotes the initial state constraint, u(0) = u 初始 Denotes the initial control constraint, x(T) = x 终止 Representing the terminal state constraint, u(T) = u 终止 Terminal control constraints.

[0038] A third aspect of this application provides a vehicle, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the program to implement the spatiotemporal motion planning method for autonomous driving compliance in urban scenarios as described in the above embodiments.

[0039] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the spatiotemporal motion planning method for autonomous driving compliance in urban scenarios as described in the above embodiments.

[0040] Therefore, this application has the following beneficial effects:

[0041] This application's embodiments can parse urban road traffic regulations and generate compliance constraint information in urban scenarios. Based on this compliance constraint information, spatiotemporal semantic corridor information for autonomous vehicles is constructed, and a compliance spatiotemporal feasible domain for autonomous vehicles is generated based on this information. Finally, based on the compliance spatiotemporal feasible domain, a safety potential field is constructed, and a trajectory solution model considering dynamic constraints is established. This enables the real-time generation of autonomous driving trajectories that strictly comply with traffic regulations while remaining dynamically feasible, safe, and smooth in complex urban scenarios, effectively solving the problem of existing technologies struggling to balance compliance, safety, and real-time performance.

[0042] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0043] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0044] Figure 1 This is a flowchart of a spatiotemporal motion planning method for autonomous driving compliance in an urban scenario, provided according to an embodiment of this application.

[0045] Figure 2 This is a block diagram of a spatiotemporal motion planning device for autonomous driving compliance in an urban scenario according to an embodiment of this application;

[0046] Figure 3 This is a structural schematic diagram of a vehicle according to an embodiment of this application. Detailed Implementation

[0047] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0048] The following description, with reference to the accompanying drawings, describes a method and apparatus for compliant spatiotemporal motion planning of autonomous driving in urban scenarios, based on embodiments of this application. Addressing the problems mentioned in the background section, this application provides a method for compliant spatiotemporal motion planning of autonomous driving in urban scenarios. This method involves parsing urban road traffic regulations to generate compliance constraint information for urban scenarios. Based on this compliance constraint information, spatiotemporal semantic corridor information for autonomous vehicles is constructed. Then, a compliant spatiotemporal feasible domain for autonomous vehicles is generated based on this spatiotemporal semantic corridor information. Finally, based on the compliant spatiotemporal feasible domain, a safe potential field is constructed, and a trajectory solving model considering dynamic constraints is established. This enables the real-time generation of autonomous driving trajectories that strictly comply with traffic regulations while remaining dynamically feasible, safe, and smooth in complex urban scenarios, effectively solving the problem of prior art's difficulty in simultaneously achieving compliance, safety, and real-time performance.

[0049] Specifically, Figure 1 This is a flowchart illustrating a spatiotemporal motion planning method for autonomous driving compliance in an urban setting, as provided in an embodiment of this application.

[0050] like Figure 1 As shown, the spatiotemporal motion planning method for autonomous driving compliance in this urban scenario includes the following steps:

[0051] In step S101, the texts of urban road traffic regulations and highway traffic regulations are obtained and parsed to generate compliance constraint information in the urban scenario.

[0052] In urban scenarios, compliance constraints include one or more combinations of speed limit regulations, safe distance constraints between vehicles, lane-changing behavior rules, right-of-way criteria, and stopping and yielding constraints for autonomous vehicles.

[0053] Specifically, in this embodiment of the application, the speed limit regulatory constraints are as follows:

[0054] The maximum speed limit on highways is 120 km / h, and the minimum speed limit is 60 km / h.

[0055] 60≤v 正常 ≤120 (1)

[0056] On roads with two lanes in the same direction, the minimum speed limit in the left lane is 100 km / h.

[0057] 100≤v 2，左 ≤120 (2)

[0058] For vehicles with three or more lanes in the same direction, the minimum speed limit for the leftmost lane is 110 km / h.

[0059] 90≤v 3，中 ≤120 (3)

[0060] If the speed limit indicated by a road sign is inconsistent with the speed limit for driving in the lane, the vehicle will obtain the speed limit indicated by the road sign through autonomous driving perception information and drive at the speed indicated by the road sign.

[0061] The safety distance constraints between vehicles are as follows:

[0062] When a motor vehicle is traveling on a highway at a speed exceeding 100 km / h, it must maintain a distance of at least 100 meters from the vehicle in front in the same lane.

[0063] 100≤d 车速大于100与前车距离 (4)

[0064] When a motor vehicle is traveling on a highway at a speed not exceeding 100 kilometers per hour, it should maintain a distance of at least 50 meters from the vehicle in front in the same lane.

[0065] 50≤d 车速小于100与前车距离 (5)

[0066] Lane changing behavior guidelines are as follows:

[0067] Where there are two or more lanes for motor vehicles traveling in the same direction, vehicles changing lanes must not impede the normal driving of vehicles in the relevant lanes, and the minimum safe distance d between vehicles must be maintained. min To satisfy:

[0068]

[0069] v f Let v be the speed of the vehicle in front, v be the speed of the vehicle itself, ρ be the reaction time, and α be the speed of the vehicle in front. max For maximum acceleration, β min Minimum braking acceleration, β max Maximum braking acceleration.

[0070] The criteria for determining right-of-way are as follows:

[0071] Right-of-way rules are typically used at intersections or special road sections, specifying the right of certain vehicles to pass under certain circumstances. For example, at an intersection without traffic lights, vehicles proceeding straight usually have the right-of-way. If vehicle A arrives at the intersection earlier than or equal to vehicle B, then vehicle A has the right-of-way.

[0072] If vehicle A arrives at the intersection at a time greater than vehicle B, then vehicle A has priority. (7)

[0073] The stop and yield constraint information for autonomous vehicles is as follows:

[0074] Stop-and-yield rules require vehicles to stop under certain circumstances and wait for pedestrians or other vehicles to pass. For example, when encountering a stop sign, a vehicle must come to a complete stop and then proceed only after confirming it is safe to do so.

[0075] If a pedestrian appears or perceives the stop sign, then v = 0 (8)

[0076] In step S102, spatiotemporal semantic corridor information of autonomous vehicles is constructed based on compliance constraint information in urban scenarios, and compliance spatiotemporal feasible domain of autonomous vehicles is constructed based on spatiotemporal semantic corridor information.

[0077] It is understood that the embodiments of this application can provide compliance input for generating spatiotemporal semantic corridor information for autonomous vehicles based on the compliance constraint information of formulas (1)-(8), and construct the spatiotemporal semantic corridor information for autonomous vehicles based on the compliance constraint information in the above embodiments as input.

[0078] The spatiotemporal semantic corridor information includes obstacle-type semantic elements and constraint-type semantic elements. The obstacle-type semantic elements include at least one of static and dynamic obstacle constraints and one of dynamic obstacle constraints. The constraint-type semantic elements include at least one of traffic light constraints, spatiotemporal semantic corridor driving distance constraints, and vehicle bounding box vertex constraints.

[0079] In this embodiment, obstacle-type semantic elements mainly include dynamic and static obstacles and traffic lights. The trajectories of dynamic and static obstacles in the time domain can be viewed as a series of static obstacles combined along a discrete time dimension, forming a spatiotemporal semantic corridor of obstacles composed of a series of steps. Traffic light semantic elements are represented in a three-dimensional spatiotemporal map as obstacles occupying a specific longitudinal position on the road at a specific time. In addition to obstacle-type semantic elements, there are also vehicle state constraints. Speed ​​limit signs on the road can be considered as speed constraints within vehicle state constraints; during lane changes, vehicles need to ensure they do not cross lane lines for too long, and this constraint can be represented as a time constraint. Specifically:

[0080] 1) Dynamic and static obstacle constraints

[0081] Static obstacles (such as walls, curbs, stationary vehicles, etc.) are usually represented by geometric shapes (such as polygons, circles). Constraints require that the planned trajectory does not overlap with the obstacle area.

[0082] For circular obstacles, the Euclidean space obstacle avoidance method is used, based on the vehicle's position p(t) = (x(t), y(t)) and the center O of the static obstacle. i radius r i Obstacle avoidance constraints can be expressed as:

[0083] ||p(t)-Oi ||2≥r i +r ego +d safe (9)

[0084] Where r ego Let d be the radius of the circumcircle. safe Let ||·||2 be the safe distance threshold, and let Euclidean distance be the Euclidean distance representation.

[0085] For a polygonal obstacle, its interior can be represented by a system of linear inequalities A using the convex hull. i p≤b i The constraints are:

[0086] A i p(t)≤b i +d safe (10)

[0087] Among them, A i This represents an m×2 matrix (where m is the number of sides of the polygon), with each row corresponding to the normal vector of one edge of the polygon pointing inwards. i This represents an m×1 vector, where each row corresponds to the offset of the edge from the origin (related to the normal vector).

[0088] For dynamic obstacles (such as other vehicles and pedestrians), the constraints of dynamic obstacles need to take into account the time dimension, and their future trajectories are usually described by motion prediction models.

[0089] For time-varying safety distance constraints, assume that the position of the dynamic obstacle j at time t is O. j (t), approximately circular in shape:

[0090] ||p(t)-O j ||2≥r j +r ego +d safe (t) (11)

[0091] Where, d safe (t) is a function of velocity v j (t) represents the increase in distance, where k is a constant. d safe (t)=d0+k·v j (t), d0 is the initial distance between the vehicle and the obstacle, v j (t) represents the velocity of the dynamic obstacle j.

[0092] 2) Traffic light constraints

[0093] A red light can be rendered as an obstacle occupying a specific vertical position and time period, during which [t] i ,t i +Tf This can be likened to a red light being a static obstacle barrier, meaning the position of a static obstacle on the lane's coordinates, which can be represented as...

[0094] P red ={P t red ,t∈[t i ,t i +T f ]} (12)

[0095] A green light indicates that there are no obstacles ahead, meaning you can continue.

[0096] 3) Spatiotemporal semantic corridor travel distance constraints

[0097] Each step of the constructed spatiotemporal semantic corridor is a three-dimensional cube built in the Frenét coordinate system. The physical properties of each step of the spatiotemporal semantic corridor include (s, d, t, l, w, h, P). parent ,P child ), where s represents the vertical coordinate s, the horizontal coordinate d, and the time coordinate t in three-dimensional space, and l, w, h represent the length, width, and height of the spatiotemporal semantic corridor steps, respectively, and represents the parent node and child node of the spatiotemporal semantic corridor steps.

[0098] In the planning process, lanes and time intervals Δt are used to generate spatiotemporal semantic corridor steps. One spatiotemporal semantic corridor step corresponds to the set of all drivable areas on the corresponding lane within time interval Δt. In the i-th time interval, i.e., [t i ,t i Therefore, the i-th spatiotemporal semantic corridor step can be represented as:

[0099]

[0100]

[0101] Among them, v 初始 It is t 初始 The initial velocity at time [time] Indicates with v 初始 With an initial speed, the autonomous vehicle travels to time t. i The minimum speed and minimum longitudinal displacement at (the start time of the i-th time interval). and The table is v initial With an initial speed, the autonomous vehicle travels to time t. i The maximum speed and maximum longitudinal displacement are calculated as follows: +Δt (the end time of the i-th time interval). min a max v min vmax These represent the minimum acceleration, maximum acceleration, minimum speed, and maximum speed of an autonomous vehicle, respectively.

[0102] 4) Vehicle bounding box vertex constraints

[0103]

[0104] Among them, P k,1 P k,2 P k,3 P k,4 Let L be the left front, right front, left rear, and right rear vertices of the vehicle's bounding box at time k. f L r These are the distances from the center of the rear axle to the front of the vehicle and from the center of the rear axle to the rear of the vehicle, respectively, with W representing the width of the vehicle.

[0105] Furthermore, embodiments of this application can construct a spatiotemporal semantic feasible domain objective function based on the above constraints:

[0106] The cost of corridor size can be expressed as:

[0107]

[0108] The cost of the corridor center point can be expressed as:

[0109]

[0110] Finally, the objective function for minimizing the compliance spatiotemporal feasible region can be expressed as:

[0111]

[0112] formulas st (1)-(20)

[0113] Where t represents time, and κ1 and κ2 represent weight coefficients. This represents the cost of the corridor size at time t. Let J represent the cost at the center point of the corridor at time t, in minimizing the objective function J. STSC In addition, the embodiments of this application also need to satisfy the constraints defined by formulas (1) to (20).

[0114] In step S103, a safety potential field for autonomous vehicles is constructed based on the compliance spatiotemporal feasible domain, and the compliance target motion trajectory of autonomous vehicles is planned based on the gradient of the safety potential field and the dynamic model of autonomous vehicles.

[0115] It is understood that the embodiments of this application can construct the safety potential field of autonomous vehicles based on the compliance spatiotemporal feasible domain constructed in the above embodiments. Specifically,

[0116] First, calculate the longitudinal direction d of the vehicle and other vehicles j at time i. x and horizontal d y Original distance value:

[0117]

[0118] Among them, O x (i,j) represents the longitudinal position of car j at time i, Ego(i) represents the longitudinal position of car j at time i, and y safe This indicates the horizontal safety threshold.

[0119] Considering the heading angle of the target obstacle vehicle The rear longitudinal D x Distance and Lateral D y Distance can be represented as

[0120]

[0121] Longitudinal safe distance x between the vehicle and other vehicles n and horizontal y n The safe distance is:

[0122]

[0123] Where X0 represents the minimum longitudinal safety distance, and Y0 represents the minimum lateral safety distance. Indicates the longitudinal speed of the vehicle. T represents the lateral velocity of the vehicle. safe This indicates the preset safety buffer time (such as braking reaction time), v j A represents the speed of car j. n This indicates the vehicle's acceleration information.

[0124] The angle of the safe potential field (the normalized orientation angle of the obstacle relative to the vehicle) can be obtained:

[0125]

[0126] Where atan2 is the arctangent function in the four quadrants, ensuring that the angle range is (-π, π], D x D y This indicates that (d) x ,d y The components are transformed to the obstacle's local coordinate system, and the obstacle's orientation is considered through a rotation matrix to align the distance calculation with the obstacle's movement direction. n Indicates the longitudinal dynamic safety distance, y n This indicates the lateral dynamic safety distance.

[0127] The normalized distance of the safe potential field (the normalized Euclidean distance from the current point to the obstacle) D safe It can be represented as:

[0128]

[0129] Therefore, the safety potential field value ρ safe It can be represented as:

[0130]

[0131] Where a1 represents the control potential field strength and b1 represents the control decay rate. The larger b1 is, the faster the potential field decays.

[0132] Next, the gradient of the safety potential field is calculated based on the safety potential field value.

[0133] The first derivative h of the safe potential field s and the second derivative h ss for:

[0134]

[0135] The partial derivatives of the normalized distance with respect to the X and Y coordinates can be expressed as:

[0136]

[0137] γ safe This represents the angle of the safe potential field.

[0138] Finally, the gradient of the potential field can be obtained:

[0139]

[0140] Furthermore, in the embodiments of this application, trajectory planning can be solved based on the obtained gradient and the established dynamic model.

[0141] First, construct matrix a as an 8×8 Jacobian matrix, whose elements correspond to the partial derivatives of the vehicle dynamics equations, which can be expressed as:

[0142]

[0143] Wherein, the state vector Control input: u = [δ f ,F x ] T

[0144]

[0145] These represent the changes in the vehicle's longitudinal velocity, lateral velocity, yaw rate, roll rate, and global coordinate system X,Yu = [δ] f ,F x ] T This indicates the front wheel steering angle and longitudinal force.

[0146] C f Front tire lateral stiffness, C r Rear tire lateral stiffness, l f The distance from the center of gravity to the front axle, l r Distance from center of gravity to rear axle, m; total vehicle mass, m²; suspension mass, I. z Moment of inertia about the Z-axis, I x Moment of inertia about the X-axis, h is the perpendicular distance from the center of rotation to the center of mass, and B is the moment of inertia about the X-axis. r Suspension roll damping, K r Suspension stiffness.

[0147] Finally, modeling the spatiotemporal motion planning for autonomous driving compliance in urban scenarios as an optimal control problem enables autonomous driving to generate compliant trajectories in complex interaction scenarios. The specific form is as follows:

[0148]

[0149] x(0)=x 初始 u(0)=u 初始

[0150] x(T)=x 终止 u(T)=u 终止

[0151] In summary, this application's embodiments transform complex traffic regulations (including speed limits, safe distances between vehicles, lane-changing rules, right-of-way determination, and stop-and-yield rules) into a computable mathematical constraint model (Equations 1-8), ensuring that the subsequently generated trajectories comply with traffic regulations from the outset. Based on compliance constraints, and comprehensively considering spatial constraints of dynamic and static obstacles, traffic light timing constraints, spatiotemporal semantic corridor boundary constraints, and vehicle collision envelope (bounding box) constraints, a compliant feasible domain that satisfies both regulatory requirements and environmental safety is constructed, accurately characterizing the drivable area of ​​the vehicle within a specific time and space. Based on the constructed compliant feasible domain, the risk between the vehicle and other vehicles (obstacles) is quantified using safety potential field theory. This method not only considers relative distance but also introduces dynamic factors such as heading angle, speed, and acceleration to calculate normalized longitudinal and lateral safe distances and construct a dynamic safety potential field (Equation 29) and its gradient (Equation 32), thereby more accurately reflecting real-time collision risks. Finally, the motion planning problem is modeled as an optimal control problem, with its objective function incorporating the influence of corridor size, center point distance, and safety potential field. During the solution process, a vehicle dynamics model (Equation 33) is introduced to ensure that the generated trajectory not only meets compliance and safety requirements but also conforms to the vehicle's own kinematics and dynamics, guaranteeing the smoothness and executability of the trajectory.

[0152] The spatiotemporal motion planning method for autonomous driving compliance in urban scenarios proposed in this application analyzes urban road traffic regulations and generates compliance constraint information for urban scenarios. Based on this compliance constraint information, it constructs spatiotemporal semantic corridor information for autonomous vehicles and generates a compliance spatiotemporal feasible domain for autonomous vehicles. Finally, based on the compliance spatiotemporal feasible domain, it constructs a safety potential field and establishes a trajectory solving model considering dynamic constraints. This enables the real-time generation of autonomous driving trajectories that strictly comply with traffic regulations while remaining dynamically feasible, safe, and smooth in complex urban scenarios, effectively solving the problem of existing technologies struggling to balance compliance, safety, and real-time performance.

[0153] Next, referring to the accompanying drawings, a spatiotemporal motion planning device for autonomous driving compliance in urban scenarios, proposed according to an embodiment of this application, is described.

[0154] Figure 2 This is a block diagram of a spatiotemporal motion planning device for autonomous driving compliance in an urban scenario, according to an embodiment of this application.

[0155] like Figure 2 As shown, the spatiotemporal motion planning device 10 for autonomous driving compliance in urban scenarios includes: an acquisition module 100, a construction module 200, and a planning module 300.

[0156] The acquisition module 100 is used to acquire and parse the texts of urban road traffic regulations and highway traffic regulations to generate compliance constraint information in urban scenarios. The construction module 200 is used to construct the spatiotemporal semantic corridor information of autonomous vehicles based on the compliance constraint information in urban scenarios, and to construct the compliance spatiotemporal feasible domain of autonomous vehicles based on the spatiotemporal semantic corridor information. The planning module 300 is used to construct the safety potential field of autonomous vehicles based on the compliance spatiotemporal feasible domain, and to plan the compliance target motion trajectory of autonomous vehicles based on the gradient of the safety potential field and the dynamic model of autonomous vehicles.

[0157] In one embodiment of this application, the compliance constraint information in urban scenarios includes one or more combinations of speed limit regulations, vehicle safety distance constraints, lane changing behavior rules, right-of-way determination criteria, and stopping and yielding constraints for autonomous vehicles.

[0158] In one embodiment of this application, the spatiotemporal semantic corridor information includes obstacle-type semantic elements and constraint-type semantic elements. The obstacle-type semantic elements include at least one of static and dynamic obstacle constraints and dynamic obstacle constraints, and the constraint-type semantic elements include at least one of traffic light constraints, spatiotemporal semantic corridor driving distance constraints, and vehicle bounding box vertex constraints.

[0159] In one embodiment of this application, the objective function of the compliance spatiotemporal feasible domain is:

[0160]

[0161] Where t represents time, and κ1 and κ2 represent weight coefficients. This represents the cost of the corridor size at time t. This represents the cost of the corridor center point at time t.

[0162] In one embodiment of this application, the first derivative h of the safety potential field is... s and the second derivative h ss for:

[0163]

[0164] Where, ρ safe D represents the safety potential field value. safe denoted by , a1 represents the normalized distance of the safety potential field, b1 represents the control potential field strength, and b1 represents the control decay rate.

[0165] In one embodiment of this application, the formula for calculating the gradient of the safety potential field is:

[0166]

[0167] Among them, among them, This represents the longitudinal component of the repulsive force (X-axis direction). S represents the partial derivative of the normalized distance with respect to the vehicle's global coordinate X, where S represents the normalized distance and X represents the X-axis coordinate. This represents the transverse (Y-axis) component of the repulsive force. This represents the partial derivative of the normalized distance with respect to the vehicle's global coordinates Y, where Y represents the Y-axis coordinate. This represents the curvature in the X direction (a measure of the degree of curvature of the potential field in pure longitudinal (X) motion. A larger value indicates a deeper or steeper longitudinal "potential well"). This represents the square of the unit direction vector in the X direction. This represents the second-order partial derivative of the normalized distance with respect to the X-coordinate. This represents the curvature in the Y direction (a measure of the degree of curvature of the potential field in pure transverse (Y) motion; the larger the value, the steeper the transverse "potential wall"). This represents the square of the unit direction vector in the Y direction. Let denote the second partial derivative of the normalized distance with respect to the Y-coordinate. This indicates the cross curvature (a measure of the coupling relationship between longitudinal and lateral motion). This represents the product of the partial derivative of the normalized distance S with respect to the X coordinate and the partial derivative of the normalized distance S with respect to the Y coordinate. This means that the normalized distance S is first derived with respect to Y, and then with respect to X (this is the mixed partial derivative, which measures the direction vector in the X direction when moving along the Y direction). How fast things are changing.

[0168] In one embodiment of this application, the formula for planning the trajectory of the compliance target is:

[0169]

[0170] x(0)=x 初始 u(0)=u 初始

[0171] x(T)=x 终止 u(T)=u 终止

[0172] in, Let J(x(t), u(t)) represent the objective function for minimizing the generated trajectory, where x(t) represents the state variable, t represents time, u(t) represents the control variable, and J(x(t), u(t)) represents the cost function. Let f(x(t),u(t)) represent the time derivative of the state variable x(t), and let f(x(t),u(t)) represent the state equation, where x(0) = x 初始 Denotes the initial state constraint, u(0) = u初始 Denotes the initial control constraint, x(T) = x 终止 Representing the terminal state constraint, u(T) = u 终止 Terminal control constraints.

[0173] It should be noted that the foregoing explanation of the embodiment of the spatiotemporal motion planning method for autonomous driving compliance in urban scenarios also applies to the spatiotemporal motion planning device for autonomous driving compliance in urban scenarios in this embodiment, and will not be repeated here.

[0174] The urban autonomous driving compliance spatiotemporal motion planning device proposed in this application analyzes urban road traffic regulations and generates compliance constraint information for urban scenarios. Based on this constraint information, it constructs spatiotemporal semantic corridor information for autonomous vehicles and generates a compliance spatiotemporal feasible domain for the autonomous vehicles. Finally, based on this compliance spatiotemporal feasible domain, it constructs a safety potential field and establishes a trajectory solving model considering dynamic constraints. This enables the real-time generation of autonomous driving trajectories that strictly comply with traffic regulations while remaining dynamically feasible, safe, and smooth in complex urban scenarios, effectively solving the problem of prior art's difficulty in simultaneously achieving compliance, safety, and real-time performance.

[0175] Figure 3 A schematic diagram of the structure of a vehicle provided in an embodiment of this application. The vehicle may include:

[0176] The memory 301, the processor 302, and the computer program stored on the memory 301 and capable of running on the processor 302.

[0177] When the processor 302 executes the program, it implements the spatiotemporal motion planning method for autonomous driving compliance in urban scenarios provided in the above embodiments.

[0178] Furthermore, the vehicle also includes:

[0179] Communication interface 303 is used for communication between memory 301 and processor 302.

[0180] The memory 301 is used to store computer programs that can run on the processor 302.

[0181] The memory 301 may include high-speed RAM (Random Access Memory) memory, and may also include non-volatile memory, such as at least one disk storage.

[0182] If the memory 301, processor 302, and communication interface 303 are implemented independently, then the communication interface 303, memory 301, and processor 302 can be interconnected via a bus to complete communication between them. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0183] Optionally, in a specific implementation, if the memory 301, processor 302, and communication interface 303 are integrated on a single chip, then the memory 301, processor 302, and communication interface 303 can communicate with each other through an internal interface.

[0184] Processor 302 may be a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement embodiments of this application.

[0185] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for compliant spatiotemporal motion planning in urban scenarios for autonomous driving.

[0186] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0187] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0188] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0189] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (FPGAs), field-programmable gate arrays (FPGAs), etc.

[0190] Those skilled in the art will understand that all or part of the steps of the methods implementing the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0191] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A spatiotemporal motion planning method for autonomous driving compliance in urban scenarios, characterized in that, Includes the following steps: Obtain the texts of urban road traffic regulations and highway traffic regulations, parse the texts of urban road traffic regulations and highway traffic regulations, and generate compliance constraint information in urban scenarios; Based on the compliance constraint information in the urban scenario, the spatiotemporal semantic corridor information of the autonomous vehicle is constructed, and the compliance spatiotemporal feasible domain of the autonomous vehicle is constructed based on the spatiotemporal semantic corridor information. The safety potential field of the autonomous vehicle is constructed based on the compliance spatiotemporal feasible domain, and the compliance target motion trajectory of the autonomous vehicle is planned based on the gradient of the safety potential field and the dynamic model of the autonomous vehicle. The first derivative of the safety potential field h s and second derivative h ss for: in, Represents the safety potential field value. Represents the normalized distance of the safety potential field. b1 represents the control potential field strength, and b1 represents the control decay rate. The formula for calculating the gradient of the safety potential field is: , in, This represents the component of the repulsive force along the X-axis. This represents the partial derivative of the normalized distance with respect to the vehicle's global coordinate X. Represents the normalized distance. This represents the component of the repulsive force along the Y-axis. This represents the partial derivative of the normalized distance with respect to the vehicle's global coordinate Y. This represents the curvature of the x-axis itself. This represents the square of the unit direction vector along the X-axis. This represents the second-order partial derivative of the normalized distance with respect to the X-coordinate. This represents the curvature itself in the Y direction. This represents the square of the unit direction vector in the Y direction. This represents the second-order partial derivative of the normalized distance with respect to the Y-coordinate. Indicates the cross curvature.

2. The spatiotemporal motion planning method for autonomous driving compliance in urban scenarios according to claim 1, characterized in that, The compliance constraints in the urban scenario include one or more combinations of speed limit regulations, safe distance constraints between vehicles, lane changing behavior regulations, right-of-way determination criteria, and stop and yield constraints for autonomous vehicles.

3. The spatiotemporal motion planning method for autonomous driving compliance in urban scenarios according to claim 1, characterized in that, The spatiotemporal semantic corridor information includes obstacle-type semantic elements and constraint-type semantic elements. The obstacle-type semantic elements include at least one of static and dynamic obstacle constraints and dynamic obstacle constraints. The constraint-type semantic elements include at least one of traffic light constraints, spatiotemporal semantic corridor driving distance constraints, and vehicle bounding box vertex constraints.

4. The spatiotemporal motion planning method for autonomous driving compliance in urban scenarios according to claim 1, characterized in that, The objective function of the compliance spatiotemporal feasible region is: in, Indicates time, and Indicates the weighting coefficient. Indicates in The cost of the corridor size at any moment Indicates in The cost of the central point of the corridor at any moment.

5. The spatiotemporal motion planning method for autonomous driving compliance in urban scenarios according to claim 1, characterized in that, The formula for planning the trajectory of the compliance target is: in, Let represent the objective function for minimizing the generated trajectory. x ( t ) represents a state variable. t Indicates time, u ( t ) represents a control variable. J ( x ( t ), u ( t )) represents the cost function. Represents state variables x ( t The derivative with respect to time, f ( x ( t ), u ( t )) represents the state equation. This represents the initial state constraints. Indicates the initial control constraints. Indicates terminal state constraints. Terminal control constraints.

6. A spatiotemporal motion planning device for autonomous driving compliance in urban scenarios, characterized in that, include: The acquisition module is used to acquire urban road traffic regulations text and highway traffic regulations text, parse the urban road traffic regulations text and the highway traffic regulations text to generate compliance constraint information in urban scenarios; The construction module is used to construct the spatiotemporal semantic corridor information of the autonomous vehicle based on the compliance constraint information in the urban scenario, and to construct the compliance spatiotemporal feasible domain of the autonomous vehicle based on the spatiotemporal semantic corridor information. The planning module is used to construct a safety potential field for the autonomous vehicle based on the compliance spatiotemporal feasible domain, and to plan the compliance target motion trajectory of the autonomous vehicle based on the gradient of the safety potential field and the dynamic model of the autonomous vehicle; wherein, The first derivative of the safety potential field h s and second derivative h ss for: in, Represents the safety potential field value. Represents the normalized distance of the safety potential field. b1 represents the control potential field strength, and b1 represents the control decay rate; where, The formula for calculating the gradient of the safety potential field is: , in, This represents the component of the repulsive force along the X-axis. This represents the partial derivative of the normalized distance with respect to the vehicle's global coordinate X. Represents the normalized distance. This represents the component of the repulsive force along the Y-axis. This represents the partial derivative of the normalized distance with respect to the vehicle's global coordinate Y. This represents the curvature of the x-axis itself. This represents the square of the unit direction vector along the X-axis. This represents the second-order partial derivative of the normalized distance with respect to the X-coordinate. This represents the curvature itself in the Y direction. This represents the square of the unit direction vector in the Y direction. This represents the second-order partial derivative of the normalized distance with respect to the Y-coordinate. Indicates the cross curvature.

7. A vehicle, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the spatiotemporal motion planning method for autonomous driving compliance in urban scenarios as described in any one of claims 1-5.

8. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed, they implement the spatiotemporal motion planning method for autonomous driving compliance in urban scenarios as described in any one of claims 1-5.

Images