Vehicle trajectory planning method and device and vehicle

Abstract

The invention discloses a vehicle trajectory planning method and device and a vehicle, and relates to the technical field of intelligent driving. The method comprises the following steps: obtaining vehicle state information of a target vehicle, a reference trajectory and environment perception data containing available traffic width, and performing trajectory optimization processing according to the vehicle state information, the reference trajectory and the environment perception data to obtain a trajectory optimization result of the reference trajectory; in the trajectory optimization processing process, taking a first passing cost required for minimizing a trajectory optimization result as a target; a passing width cost item in the first passing cost is a power function taking the residual passing width as an independent variable, and the passing width cost item is increased along with the reduction of the residual passing width; the remaining passing width is the difference between the available passing width and the vehicle width of the target vehicle. Therefore, in a narrow traffic scene, the traffic width cost item is increased in a super-linear manner, the constraint strength on the transverse position of the track is enhanced, the generation of a high-risk track excessively close to an obstacle is effectively avoided, the planning interruption frequency is reduced, and the trafficability of the vehicle in the narrow scene is remarkably improved.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, specifically to a vehicle trajectory planning method, device, and vehicle. Background Technology

[0002] Vehicle trajectory planning has become one of the core components for realizing intelligent driving functions. In a hierarchical planning system, a coarse reference trajectory from the starting point to the end point is usually generated by the global planning layer first, and then the local planning layer makes fine-grained adjustments based on real-time environmental information to generate a local trajectory that can be directly tracked and executed.

[0003] Currently, existing local trajectory optimization algorithms can basically meet the requirements in conventional wide road scenarios, but their passability is poor in narrow road scenarios such as internal roads in residential areas and narrow rural roads. Summary of the Invention

[0004] In view of this, this application aims to provide a vehicle trajectory planning method, device and vehicle that can improve the vehicle's passability in narrow scenarios.

[0005] According to a first aspect of this application, a vehicle trajectory planning method is provided, comprising: acquiring vehicle state information of a target vehicle, a reference trajectory, and environmental perception data of the environment in which the target vehicle is located; the environmental perception data including available passage widths for characterizing the minimum distance between each location in the environment and surrounding obstacles; performing trajectory optimization processing based on the vehicle state information, the reference trajectory, and the environmental perception data to obtain a trajectory optimization result of the reference trajectory; in the trajectory optimization processing, minimizing a first passage cost required for the trajectory optimization result is the objective; the first passage cost includes a passage width cost term; the passage width cost term is a power function with the remaining passage width as the independent variable, and the passage width cost term increases as the remaining passage width decreases; the remaining passage width is the difference between the available passage width and the vehicle width of the target vehicle. By acquiring the vehicle status information, reference trajectory, and environmental perception data including available passage width of the target vehicle, trajectory optimization is performed with the goal of minimizing the first passage cost required for the trajectory optimization result, resulting in the trajectory optimization result of the reference trajectory. The first passage cost includes a passage width cost term, which is a power function with the remaining passage width as the independent variable. Because the passage width cost term uses a power function design, its value increases rapidly and non-linearly as the remaining passage width gradually decreases. This creates a constraint effect far stronger than traditional linear costs in the critical region of obstacles, effectively avoiding situations where the vehicle is allowed to get too close to obstacles to improve trajectory tracking accuracy. It also reduces the phenomenon of no feasible solution to the optimization problem and interruption of the planning process due to triggering the minimum safety margin constraint, ultimately significantly improving the passability of the target vehicle in scenarios with limited passage space, such as narrow roads. When the road width is sufficient and the vehicle is far from obstacles, the remaining passage width increases significantly, and the passage width cost term decreases substantially. At this time, other terms in the first passage cost dominate, allowing the vehicle to focus on trajectory tracking accuracy and driving smoothness as the main optimization objectives. This allows for a reasonable optimization focus across different traffic width scenarios. This not only significantly improves the vehicle's maneuverability in confined spaces such as narrow roads, but also ensures trajectory tracking accuracy and driving smoothness on open roads, thereby enhancing the autonomous driving system's environmental adaptability and overall performance under various operating conditions.

[0006] Optionally, trajectory optimization processing is performed based on the vehicle status information, the reference trajectory, and the environmental perception data to obtain the trajectory optimization result of the reference trajectory. This includes: solving the objective function of the local optimization algorithm to obtain the control parameters of the target vehicle; the objective function represents the first passage cost required for the trajectory optimization result; the value of the objective function is calculated based on the vehicle status information, the reference trajectory, and the environmental perception data; the objective function includes a passage width cost term; when the remaining passage width exceeds the minimum safety margin, the passage width cost term is a power function with the remaining passage width as the independent variable. By acquiring the vehicle status information of the target vehicle, the reference trajectory, and environmental perception data containing the available passage width, the control parameters are solved with the objective function of the local optimization algorithm as the objective, wherein the objective function includes a passage width cost term, and when the remaining passage width exceeds the minimum safety margin, this cost term is a power function with the remaining passage width as the independent variable. Because the passage width cost term uses a power function design, its value increases rapidly and non-linearly as the remaining passage width gradually decreases. This creates a constraint effect far stronger than traditional linear cost in the critical region of obstacles, effectively preventing local optimization algorithms from allowing vehicles to get too close to obstacles in order to maintain trajectory tracking accuracy. It also reduces the occurrence of infeasible solutions and planning interruptions caused by triggering minimum safety margin constraints, ultimately significantly improving the vehicle's passability in scenarios with limited passage space, such as narrow roads. Conversely, when the road width is sufficient and the vehicle is far from obstacles, the remaining passage width increases significantly, and the passage width cost term decreases substantially. At this point, other terms in the objective function become dominant, allowing the vehicle to focus on trajectory tracking accuracy and driving smoothness as the primary optimization objectives. This maintains a reasonable optimization focus under different passage width scenarios. This not only significantly improves the vehicle's passability in scenarios with limited passage space, such as narrow roads, but also balances trajectory tracking accuracy and driving smoothness in open road sections, thereby enhancing the autonomous driving system's environmental adaptability and overall performance under different operating conditions.

[0007] Optionally, obtaining the reference trajectory of the target vehicle includes: obtaining vehicle state information, target pose, and environmental perception data of the environment in which the target vehicle is located; wherein, the vehicle state information includes the initial pose of the target vehicle, and the environmental perception data includes information for characterizing the distribution of obstacles and the range of passable areas in the environment; determining the centerline of the passable area in the environment in which the target vehicle is located based on the environmental perception data; performing trajectory planning based on the vehicle state information, the target pose, the environmental perception data, and the centerline to generate a reference trajectory connecting the initial pose and the target pose; during the trajectory planning process, minimizing the second passage cost required for the reference trajectory is the objective; wherein, the second passage cost includes a centerline deviation cost; the centerline deviation cost is positively correlated with the deviation distance of the target vehicle from the centerline of the passable area at each node in the planned path.

[0008] Optionally, the step of generating a reference trajectory connecting the initial pose and the target pose by performing trajectory planning based on the vehicle state information, the target pose, the environmental perception data, and the centerline includes: generating the reference trajectory based on a global planning algorithm with the objective of minimizing the second passage cost; wherein the second passage cost includes actual cost and heuristic cost, the actual cost being the cumulative travel cost of each node from the initial pose to the target pose, and the heuristic cost including the centerline deviation cost. By acquiring the target vehicle's state information, target pose, and environmental perception data including obstacle distribution and the range of the passable area, after determining the centerline of the passable area, the reference trajectory is generated with the objective of minimizing the total cost of the global planning algorithm. The total cost includes the centerline deviation cost, which is positively correlated with the deviation distance. This allows the global planning algorithm to actively guide each node of the reference trajectory to be as close as possible to the centerline of the passable area during optimization, due to the characteristic that the centerline deviation cost increases with the deviation distance, thus helping to improve vehicle passage safety.

[0009] Optionally, the centerline deviation cost is the product of the deviation distance and the weight value; the weight value is negatively correlated with the available passage width. By designing the centerline deviation cost as the product of the deviation distance and the weight value, and maintaining a negative correlation between the weight value and the available passage width, dynamic adaptive adjustment of the centerline deviation cost is achieved. In narrow road sections, the available passage width is smaller, and the weight value increases accordingly, significantly increasing the centerline deviation cost. This generates a stronger centripetal force during path optimization, forcing the generated reference trajectory to travel closer to the centerline of the passable area, effectively avoiding the problem of decreased traffic safety due to excessive deviation from the centerline. In wide road sections, the available passage width is larger, and the weight value decreases accordingly, reducing the centerline deviation cost and allowing path planning to focus more on other optimization objectives. This mechanism of dynamically adjusting the centerline deviation cost based on the passage width can automatically balance path centering requirements with other planning objectives in road environments of different widths. It ensures driving safety in narrow road sections while maintaining planning flexibility in wide road sections, significantly improving the adaptability and practicality of the global trajectory planning algorithm in different scenarios.

[0010] Optionally, the weight value is a continuous piecewise function with the available passage width as the independent variable; when the available passage width is less than a preset narrow threshold, the weight value reaches its maximum value; when the available passage width is greater than a preset wide threshold, the weight value reaches its minimum value; when the available passage width is between the narrow threshold and the wide threshold, the weight value decreases linearly from the maximum value to the minimum value as the available passage width increases. By designing the weight value as a continuous piecewise function with the available passage width as the independent variable, precise control of the centerline deviation cost intensity is achieved. When the available passage width is below the narrowness threshold, the weight value remains at its maximum. In this case, the centerline deviation cost dominates the total cost, forcing the planned path to closely follow the centerline of the passable area, maximizing the safety of passage in narrow sections. When the available passage width is above the wideness threshold, the weight value drops to its minimum, and the impact of the centerline deviation cost is correspondingly weakened. The algorithm can then prioritize optimizing other performance indicators such as path length and smoothness. When the available passage width is in the intermediate transition range, the weight value decreases linearly with the increase in width, allowing the planning strategy to smoothly transition from "safety first" to "efficiency first." This continuously segmented weight design avoids abrupt changes in planning results caused by sudden weight changes and ensures that a reference trajectory that best meets the safety requirements of the current passage conditions can be generated under different width scenarios, significantly improving the adaptability and reliability of the global path planning system.

[0011] Optionally, the weight value is an S-shaped function with the available passage width as the independent variable, and the weight value smoothly decreases from its maximum value to its minimum value as the available passage width increases. By using an S-shaped function as the calculation method for the weight value, the weight value exhibits a smooth, non-linear decreasing characteristic as the available passage width increases, achieving continuous and natural adjustment of the centerline deviation cost intensity. This S-shaped function maintains a high weight value near the narrow threshold, ensuring sufficient centerline deviation penalty force when the passage width is limited, forcing the planned path to closely adhere to the centerline of the passable area to ensure passage safety; near the wide threshold, the weight value gradually converges to its minimum value, weakening the constraint on path centering and allowing the algorithm to prioritize optimizing other performance indicators such as path length and smoothness; in the intermediate transition region, the smoothness of the S-shaped function avoids the jitter in the planned trajectory that may be caused by sudden weight changes, ensuring the stability of the generated reference trajectory. This weight adjustment mechanism based on the sigmoid function enables the global planning algorithm to automatically adjust its optimization strategy according to changes in the passage width. While ensuring safe passage in narrow areas, it fully leverages the planning flexibility in wide areas, significantly improving the adaptability and robustness of the path planning system in different scenarios.

[0012] Optionally, obtaining the reference trajectory of the target vehicle includes: acquiring historical trajectory information recorded by the target vehicle during its historical driving process; and performing reverse time-series processing on the historical trajectory information to obtain the reference trajectory. Generating the reference trajectory using reverse time-series processing of historical trajectory information enables the target vehicle to plan based on the actual driving path during line-following reversing, improving the accuracy and reliability of the trajectory. Simultaneously, by introducing a passage width cost term into a local optimization algorithm that aims to minimize the objective function, and dynamically adjusting the cost intensity in the form of a power function of the remaining passage width—increasing the passage width cost term when the remaining passage width decreases—obstacle avoidance constraints in narrow areas are strengthened during reversing, ensuring the vehicle maintains a safety margin during line-following reversing and improving driving safety and environmental adaptability.

[0013] Optionally, before performing trajectory optimization processing based on the vehicle state information, the reference trajectory, and the environmental perception data to obtain the trajectory optimization result of the reference trajectory, the method further includes: smoothing the reference trajectory to obtain a smoothed reference trajectory; the smoothed reference trajectory is a trajectory with continuous curvature. By smoothing the reference trajectory to obtain a trajectory with continuous curvature, the problem of possible abrupt curvature changes in the original trajectory is effectively solved. Curvature continuity ensures smooth and consistent steering wheel operation, avoiding steering wheel vibration caused by discontinuous trajectory curvature, and significantly improving ride comfort. Simultaneously, the smoothed reference trajectory provides input to the local optimization algorithm that is more consistent with vehicle kinematics, making the solved control parameters smoother and more reasonable, and improving the accuracy and stability of trajectory tracking. This preprocessing mechanism effectively reduces the complexity of subsequent optimization problems and helps improve the performance and reliability of the entire trajectory planning system.

[0014] Optionally, the global planning algorithm is a hybrid A* algorithm. When generating the reference trajectory using the global planning algorithm, a hybrid A* algorithm is employed. This algorithm constructs a kinematically feasible path that satisfies the vehicle's minimum turning radius constraint within the traversable area. Simultaneously, it guides the path closer to the centerline in narrow scenarios using a centerline deviation cost term, providing high-quality initial trajectory input for local optimization and significantly reducing the risk of the vehicle scraping against obstacles.

[0015] Optionally, the local optimization algorithm is a model predictive control algorithm. Employing the MPC algorithm as the local optimization algorithm, leveraging its rolling optimization characteristics, it can solve for the objective function containing a power-law-type traffic width cost term based on real-time updated vehicle state, reference trajectory, and environmental perception data within each control cycle, generating the optimal control parameter sequence in real time. This algorithm incorporates the power-law-type traffic width cost term into the cost evaluation system within the prediction time domain. This allows the predictive model to anticipate the superlinear growth trend of the traffic width cost term when the remaining traffic width (the difference between available traffic width and vehicle width) shrinks in narrow road sections, thus proactively generating an obstacle avoidance trajectory. This combination of MPC and an objective function containing a sensitive traffic width cost term fully leverages the advantages of MPC in handling various constraints and dynamic optimization, while strengthening the precise control of the distance between the vehicle and obstacles through cost term design. Ultimately, it significantly improves the safety, real-time performance, and tracking accuracy of local trajectory planning, providing a reliable guarantee for the smooth and efficient driving of the target vehicle in various environments.

[0016] According to a second aspect of this application, a vehicle trajectory planning device is provided, comprising: The acquisition module is used to acquire the vehicle status information and reference trajectory of the target vehicle, as well as the environmental perception data of the environment in which the target vehicle is located; the environmental perception data includes the available passage width used to characterize the minimum distance between each location in the environment and its surrounding obstacles. The local optimization module is used to solve for the objective function of the local optimization algorithm to obtain the control parameters of the target vehicle. The value of the objective function is calculated based on the vehicle state information, the reference trajectory, and the environmental perception data. The objective function includes a traffic width cost term. When the remaining traffic width exceeds the minimum safety margin, the traffic width cost term is a power function with the remaining traffic width as the independent variable, and the traffic width cost term increases as the remaining traffic width decreases. The remaining traffic width is the difference between the available traffic width and the vehicle width of the target vehicle.

[0017] According to a third aspect of this application, a computer-readable storage medium is provided, the storage medium storing a computer program for performing the methods described in any of the above embodiments.

[0018] According to a fourth aspect of this application, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; the processor being configured to perform the method described in any of the above embodiments.

[0019] According to a fifth aspect of this application, a vehicle is provided, including the aforementioned electronic equipment.

[0020] This application provides a vehicle trajectory planning method, device, and vehicle. The method acquires vehicle state information of the target vehicle, a reference trajectory, and environmental perception data including available passage width. Based on this, trajectory optimization is performed to obtain the optimized trajectory result. During trajectory optimization, the objective is to minimize the first passage cost required to achieve the optimized trajectory result. The passage width cost term in the first passage cost is a power function with the remaining passage width as the independent variable, and the passage width cost term increases as the remaining passage width decreases. The remaining passage width is the difference between the available passage width and the target vehicle's width. Therefore, in narrow passage scenarios, the passage width cost term grows superlinearly, strengthening the constraint on the lateral position of the trajectory, effectively avoiding the generation of high-risk trajectories that are too close to obstacles, reducing the frequency of planning interruptions, and thus significantly improving the vehicle's passability in narrow scenarios. Attached Figure Description

[0021] Figure 1 The diagram shown is a flowchart of a vehicle trajectory planning method provided in one embodiment of this application.

[0022] Figure 2The diagram shown is a block diagram of a vehicle trajectory planning device provided in one embodiment of this application.

[0023] Figure 3 The diagram shown is a structural block diagram of an electronic device provided in one embodiment of this application. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Application Overview In the field of intelligent driving technology, local trajectory optimization algorithms, as a key module connecting the global reference trajectory and the actual vehicle control, need to determine control parameters such as steering angle and acceleration by solving the objective function based on the real-time status information of the target vehicle such as position, heading angle, and speed, the reference trajectory generated by global planning, and environmental perception data, so as to realize the vehicle's tracking and obstacle avoidance of the reference trajectory.

[0026] Typical methods such as model predictive control usually use cost maps to quantify environmental information. In the obstacle layer, an expansion region with gradually increasing cost is set around the obstacle. The cost in this region increases linearly as the distance from the obstacle decreases, so as to guide the planned trajectory away from the obstacle.

[0027] However, existing cost growth methods are mostly based on linear or near-linear gradient patterns, which are significantly insufficient in narrow traffic scenarios. Because the gradient of a linear cost function changes relatively smoothly, it cannot create sufficient constraint strength in the critical region near obstacles. This leads optimization algorithms to allow vehicles to approach obstacles excessively in order to maintain trajectory tracking accuracy. This approximation not only increases the actual driving risk but also easily triggers the system's minimum safe distance constraint, rendering the optimization problem unsolvable and interrupting the planning process. Consequently, it severely impacts the vehicle's passability in extreme scenarios and the system's continuous operation capability.

[0028] To address the aforementioned issues, this application introduces a power-law-based traffic width cost term when constructing the objective function of the local optimization algorithm, enhancing trajectory constraint capabilities within narrow areas. The remaining traffic width is defined as the difference between the available traffic width and the vehicle width. The traffic width cost term is a power-law function with the remaining traffic width as its independent variable, and it increases as the remaining traffic width decreases. By minimizing this objective function to obtain the vehicle's control parameters, the optimization algorithm can effectively suppress the generation of high-risk trajectories that closely follow obstacles, reducing planning interruptions caused by triggering safety constraints, thereby improving the vehicle's passability and system operational continuity in narrow traffic scenarios.

[0029] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0030] Exemplary methods Figure 1 This is a flowchart illustrating a vehicle trajectory planning method provided in one embodiment of this application. Figure 1 The method described is executed by a computing device located in the vehicle, or a server-side computing device communicatively connected to the vehicle, such as an onboard controller, domain controller, or cloud server; this application embodiment does not limit this. Figure 1 As shown, the method includes the following: Step S110: Obtain the vehicle status information and reference trajectory of the target vehicle, as well as the environmental perception data of the environment in which the target vehicle is located; the environmental perception data includes the available passage width used to characterize the minimum distance between each location in the environment and its surrounding obstacles.

[0031] In this embodiment of the application, the vehicle status information is used to reflect the vehicle's operating status in real time, and may include dynamic parameters such as the vehicle's position, speed, acceleration, and orientation angle.

[0032] In this embodiment of the application, the reference trajectory may refer to the expected driving path of the target vehicle.

[0033] In this embodiment of the application, the environmental perception data may refer to the target vehicle's perception data of the surrounding environment. The environmental perception data is acquired by vehicle-mounted perception sensors such as lidar, cameras, and millimeter-wave radar, and is used to describe the spatial distribution of obstacles around the vehicle.

[0034] In this embodiment of the application, the available passage width is the minimum distance between each location in the environment and its surrounding obstacles, which is used to quantify the safe passage space of that location. Its value is stored in a map layer in the cost map that is specifically used to characterize the passage space.

[0035] Step S120: Perform trajectory optimization processing based on the vehicle status information, the reference trajectory, and the environmental perception data to obtain the trajectory optimization result of the reference trajectory; during the trajectory optimization process, the objective is to minimize the first traffic cost required for the trajectory optimization result; the first traffic cost includes a traffic width cost term; the traffic width cost term is a power function with the remaining traffic width as the independent variable, and the traffic width cost term increases as the remaining traffic width decreases; the remaining traffic width is the difference between the available traffic width and the vehicle width of the target vehicle.

[0036] In this embodiment, the trajectory optimization result is obtained by optimizing a reference trajectory. This optimization result can be a global or local trajectory of the reference trajectory, or it can be control parameters used to control the vehicle. The trajectory optimization result generated through the trajectory optimization process must satisfy vehicle dynamics constraints and environmental traffic constraints.

[0037] In this embodiment of the application, the first passage cost is used to represent the cost of the candidate trajectory in the trajectory optimization process, including a passage width cost term and may include other cost terms.

[0038] In this embodiment of the application, the remaining passage width is the difference between the available passage width and the vehicle width of the target vehicle, that is, the actual safe space for vehicles to pass through at a certain location in the environment, and its value can reflect the degree of passage safety at that location.

[0039] In this embodiment, the passage width cost term is a power function with the remaining passage width as the independent variable. The passage width cost term increases rapidly as the remaining passage width decreases; when the available passage width is large, the passage width cost term increases as the remaining passage width decreases. The calculation formula for the passage width cost term can be: In the formula, Width represents the available passage width at a certain location in the environment, which is the minimum distance between that location and surrounding obstacles; This indicates the width of the target vehicle; is the remaining passage width at this location; K is a preset scaling factor used to control the overall weight of the passage width cost term; n is a preset power parameter used to control the growth rate of the power function of the cost term.

[0040] In this embodiment, by acquiring the vehicle state information of the target vehicle, the reference trajectory, and environmental perception data including the available passage width, trajectory optimization is performed with the goal of minimizing the first passage cost required for the trajectory optimization result, thus obtaining the trajectory optimization result of the reference trajectory. The first passage cost includes a passage width cost term, which is a power function with the remaining passage width as the independent variable. Because the passage width cost term uses a power function design, as the remaining passage width gradually decreases, the value of the passage width cost term increases non-linearly and rapidly. This creates a constraint effect much stronger than traditional linear costs in the critical region of obstacles, effectively avoiding situations where the vehicle is allowed to get too close to obstacles to improve trajectory tracking accuracy. It also reduces the phenomenon of no feasible solution to the optimization problem and interruption of the planning process due to triggering the minimum safety margin constraint, ultimately significantly improving the passability of the target vehicle in scenarios with limited passage space, such as narrow roads. When the road width is sufficient and the vehicle is far from obstacles, the remaining passage width increases significantly, and the passage width cost term decreases substantially. At this time, other terms in the first passage cost dominate, allowing the vehicle to focus on trajectory tracking accuracy and driving smoothness as the main optimization objectives. This allows for a reasonable optimization focus across different traffic width scenarios. This not only significantly improves the vehicle's maneuverability in confined spaces such as narrow roads, but also ensures trajectory tracking accuracy and driving smoothness on open roads, thereby enhancing the autonomous driving system's environmental adaptability and overall performance under various operating conditions.

[0041] based on Figure 1 In addition to the method described in the embodiments of this specification, some specific implementation schemes of the method are also provided, which will be described below.

[0042] Optionally, trajectory optimization processing is performed based on the vehicle status information, the reference trajectory, and the environmental perception data to obtain the trajectory optimization result of the reference trajectory, including: The objective function of the local optimization algorithm is minimized to obtain the control parameters of the target vehicle. The objective function represents the first passage cost required for the trajectory optimization result. The value of the objective function is calculated based on the vehicle state information, the reference trajectory, and the environmental perception data. The objective function includes a traffic width cost term; when the remaining traffic width exceeds the minimum safety margin, the traffic width cost term is a power function with the remaining traffic width as the independent variable.

[0043] In the embodiments of this application, the local optimization algorithm is used to solve the optimal control sequence that satisfies vehicle dynamics constraints within a finite time domain. The local optimization algorithm may include algorithms such as Model Predictive Control (MPC) and Linear Quadratic Regulator (LQR).

[0044] In this embodiment, the objective function is a mathematical expression used to evaluate the performance of control parameters, including multiple cost terms, including a traffic width cost term. Specifically, the objective function may include a state tracking error term, a control input penalty term, a control increment penalty term, etc.; wherein, the state tracking error term is used to quantify the deviation between the actual state and the desired state of the vehicle, such as tracking errors in position, speed, or heading angle; the control input penalty term is used to limit the magnitude of control commands to improve comfort, such as the absolute value of steering angle or acceleration; the control increment penalty term is used to suppress abrupt changes in control commands, such as the rate of change of acceleration or steering angle, to improve ride comfort and stability. If the objective function includes an obstacle avoidance penalty term for environmental obstacle avoidance, the traffic width cost term can be a sub-term of that obstacle avoidance penalty term.

[0045] In this embodiment of the application, the value of the objective function is calculated based on the vehicle state information, the reference trajectory, and the environmental perception data.

[0046] In this embodiment of the application, the control parameters include, but are not limited to, steering wheel angle, speed command, etc., which directly act on the vehicle's actuators to achieve trajectory tracking.

[0047] In this embodiment of the application, the minimum safety margin is a preset threshold to ensure the safe driving of the vehicle, representing the minimum safe distance that the vehicle and the obstacle must maintain. When the remaining passage width is less than or equal to this value, it indicates that the point is prohibited from being passed.

[0048] In this embodiment, by acquiring the vehicle state information, reference trajectory, and environmental perception data including available passage width of the target vehicle, the objective function value of the local optimization algorithm is minimized to solve the control parameters. The objective function includes a passage width cost term, and when the remaining passage width exceeds the minimum safety margin, this cost term is a power function with the remaining passage width as the independent variable. Since the passage width cost term adopts a power function design, as the remaining passage width gradually decreases, the value of the cost term will increase non-linearly and rapidly, forming a constraint effect much stronger than that of traditional linear cost in the critical region of obstacles. This effectively avoids the situation where the local optimization algorithm allows the vehicle to get too close to the obstacle in order to balance trajectory tracking accuracy, and reduces the phenomenon of no feasible solution to the optimization problem and interruption of the planning process caused by triggering the minimum safety margin constraint. Ultimately, it significantly improves the passability of the target vehicle in scenarios with limited passage space such as narrow roads. When the road width is sufficient and the vehicle is far away from the obstacle, the remaining passage width increases significantly, and the passage width cost term decreases significantly. At this time, other terms in the objective function take the lead, so that the vehicle focuses on trajectory tracking accuracy and driving smoothness as the main optimization objectives. This allows for a reasonable optimization focus across different traffic width scenarios. This not only significantly improves the vehicle's maneuverability in confined spaces such as narrow roads, but also ensures trajectory tracking accuracy and driving smoothness on open roads, thereby enhancing the autonomous driving system's environmental adaptability and overall performance under various operating conditions.

[0049] Optionally, obtain the reference trajectory of the target vehicle, including: The vehicle status information, target pose, and environmental perception data of the target vehicle's environment are acquired; wherein, the vehicle status information includes the target vehicle's initial pose, and the environmental perception data includes information characterizing the distribution of obstacles and the range of passable areas in the environment. Based on the environmental perception data, the centerline of the passable area in the environment where the target vehicle is located is determined; Based on the vehicle status information, the target pose, the environmental perception data, and the centerline, trajectory planning is performed to generate a reference trajectory connecting the starting pose and the target pose; during the trajectory planning process, the objective is to minimize the second passage cost required for the reference trajectory. The second passage cost includes a centerline deviation cost; the centerline deviation cost is positively correlated with the deviation distance of the target vehicle from the centerline of the passable area at each node in the planned path.

[0050] In the embodiments of this application, the vehicle state information, in addition to including the initial pose of the target vehicle, typically also includes parameters reflecting the current motion state of the vehicle, such as the real-time driving speed, acceleration, and vehicle body attitude angle.

[0051] In this embodiment of the application, the initial pose refers to the initial state of the target vehicle when it begins global path planning, which may include the vehicle's current position and heading angle, etc.

[0052] In this embodiment of the application, the target pose refers to the final state that the target vehicle is expected to reach, including the target position coordinates and the target heading angle.

[0053] In the embodiments of this application, the environmental perception data may refer to the raw or processed information of the environment around the vehicle, mainly including the position, size, shape, speed and direction of obstacles, as well as the boundary outline and topology of the passable area.

[0054] In this embodiment of the application, the passable area refers to the area in the environment that is identified as a safe and legal area for vehicles to drive in. It is usually divided based on factors such as obstacle distribution, road signs, and traffic rules in environmental perception data, such as the area within a lane, an accessible open space, or a passable road surface.

[0055] In this embodiment, the passable area refers to an area in the environment that is identified as a safe and legal area for vehicles to drive in, such as a lane area, an accessible open space, or a road surface that is permitted to pass. The passable area can typically be defined based on factors such as obstacle distribution, road signs, and traffic rules in environmental perception data.

[0056] In this embodiment, the centerline refers to an ideal reference line extracted from the traversable area using an algorithm, and is typically located at the geometric center of the traversable area. The centerline can be extracted from the boundary information of the traversable area using algorithms such as Voronoi diagrams.

[0057] In this embodiment of the application, the second passage cost is used to represent the cost required for the reference trajectory; the second passage cost includes the centerline deviation cost.

[0058] In this embodiment of the application, the deviation distance refers to the vertical distance from the current candidate node to the center line of the passable area in the path planning. This distance directly reflects the degree of lateral deviation of the path at that node relative to the ideal center position.

[0059] In this embodiment of the application, the centerline deviation cost represents a quantitative assessment of the degree to which the path planning results maintain the centerline of the passable area. It is positively correlated with the deviation distance of the current candidate node, and its specific functional relationship can be expressed as a linear function, a quadratic function, or other monotonically increasing functions.

[0060] In this embodiment, by acquiring vehicle state information, target pose, and environmental perception data including obstacle distribution and passable area range, the centerline of the passable area is determined. Then, a reference trajectory is generated with the goal of minimizing the second travel cost required for the reference trajectory. The second travel cost includes a centerline deviation cost that is positively correlated with the deviation distance. This allows the global planning algorithm to actively guide each node of the reference trajectory to be as close as possible to the centerline of the passable area during optimization, due to the characteristic that the centerline deviation cost increases with the deviation distance, thus helping to improve vehicle passage safety.

[0061] Optionally, the step of generating a reference trajectory connecting the initial pose and the target pose by performing trajectory planning based on the vehicle state information, the target pose, the environmental perception data, and the centerline includes: The reference trajectory is generated based on a global planning algorithm with the objective of minimizing the second passage cost. The second passage cost includes actual cost and heuristic cost. The actual cost is the cumulative travel cost of each node from the starting pose to the target pose, and the heuristic cost includes the centerline deviation cost.

[0062] In this embodiment of the application, the global planning algorithm is used to generate a reference trajectory connecting the starting pose and the target pose within a passable area. The global planning algorithm is not limited to the hybrid A* algorithm, but can also be a fast exploratory random tree algorithm, a probabilistic road map algorithm, or a state grid search algorithm, etc.

[0063] In this embodiment of the application, the second passage cost is the total cost of the global planning algorithm, which can refer to the path quality index used by the global planning algorithm to comprehensively evaluate the path during the path search process. It can be obtained by directly summing or weighted summing the actual cost and the heuristic cost.

[0064] In this embodiment, the actual cost refers to the cumulative travel cost from the initial pose to the current path node. Generally, the actual cost may include geometric length cost, curvature penalty cost, curvature change penalty cost, reversing distance penalty cost, gear shift number penalty cost, and obstacle gap penalty cost, etc. The geometric length cost term, representing the total geometric length of a path segment, is calculated by summing the lengths of each path segment. It is typically used as a base cost term and weighted with a fixed weight. The curvature penalty cost term penalizes sections of the path with excessive curvature to prevent sharp turns. It is calculated by weighted summing of the curvature values ​​of each path point. The curvature change penalty cost term penalizes abrupt changes in curvature between adjacent path points to prevent sudden steering wheel movements. It is achieved by weighted summing of the curvature differences between adjacent points. The reversing distance penalty cost term specifically penalizes road sections requiring reversing. It is achieved by accumulating the path length of the reversing section and applying a higher weight. The gear shift frequency penalty cost term reduces the frequency of gear shifts during driving. Its cost is proportional to the number of gear shifts. The obstacle gap penalty cost term keeps the path away from obstacles. It is achieved by weighted summing the ratio of the length of each path segment to the distance from that segment to the nearest obstacle. The closer the distance to the obstacle, the higher the value.

[0065] In this embodiment, the heuristic cost refers to the function used in the global planning algorithm to estimate the expected cost from the current node to the target pose. The heuristic cost can be composed of a basic heuristic cost and an adaptively weighted heading deviation cost. The basic heuristic cost can be determined by taking the larger value between the incomplete, barrier-free heuristic cost and the complete, barrier-free heuristic cost, i.e. The non-integrity obstacle-free heuristic cost ConstrainedCost considers the vehicle's kinematic constraints but ignores environmental obstacles during calculation, estimating the path cost by constructing a path that meets constraints such as the vehicle's minimum turning radius; the integrity obstacle-prone heuristic cost UnconstrainedCost considers environmental obstacles but temporarily ignores the vehicle's kinematic constraints, using a shortest path search algorithm on a two-dimensional grid map to estimate the obstacle avoidance path cost.

[0066] In this embodiment, by acquiring vehicle state information, target pose, and environmental perception data including obstacle distribution and passable area range, the centerline of the passable area is determined. Then, a reference trajectory is generated with the goal of minimizing the total cost of the global planning algorithm. The total cost includes the centerline deviation cost, which is positively correlated with the deviation distance. This ensures that during optimization, the global planning algorithm actively guides each node of the reference trajectory to be as close as possible to the centerline of the passable area due to the characteristic that the centerline deviation cost increases with the deviation distance, thus improving vehicle passage safety.

[0067] Optionally, the centerline deviation cost is the product of the deviation distance and the weight value; the weight value is negatively correlated with the available passage width.

[0068] In this embodiment of the application, the weight value is dynamically adjusted according to the available passage width of the passable area at the current node. The smaller the available passage width, the larger the weight value, and the larger the available passage width, the smaller the weight value.

[0069] In this embodiment of the application, the heuristic cost can be expressed as: In the formula, This indicates the deviation distance of the current candidate node from the center line of the passable area; This represents the weight value; This indicates the cost of deviation from the centerline.

[0070] In this embodiment, the centerline deviation cost is designed as the product of the deviation distance and the weight value, and the weight value is negatively correlated with the available traffic width, thus achieving dynamic adaptive adjustment of the centerline deviation cost. In narrow road sections, the available traffic width is smaller, and the weight value increases accordingly, significantly increasing the centerline deviation cost. This generates a stronger centripetal force during path optimization, forcing the generated reference trajectory to travel closer to the centerline of the passable area, effectively avoiding the decrease in traffic safety caused by excessive deviation from the centerline. In wide road sections, the available traffic width is larger, and the weight value decreases accordingly, reducing the centerline deviation cost and allowing path planning to focus more on other optimization objectives. This mechanism of dynamically adjusting the centerline deviation cost based on traffic width can automatically balance path centering requirements with other planning objectives in road environments of different widths. It ensures driving safety in narrow road sections while maintaining planning flexibility in wide road sections, significantly improving the adaptability and practicality of the global trajectory planning algorithm in different scenarios.

[0071] Optionally, the weight value is a continuous piecewise function with the available passage width as the independent variable; When the available passage width is less than a preset narrowing threshold, the weight value reaches its maximum value; When the available passage width is greater than a preset width threshold, the weight value reaches its minimum value; When the available passage width is between the narrow threshold and the wide threshold, the weight value decreases linearly from the maximum value to the minimum value as the available passage width increases.

[0072] In this embodiment of the application, the narrow threshold refers to the upper limit of the width used to determine whether the passable area belongs to a narrow road segment. When the available passable width is lower than this threshold, the system will activate the maximum weight value to strengthen the requirement for path centering.

[0073] In this embodiment of the application, the width threshold refers to the lower limit of the width used to determine whether the passable area belongs to a wide road segment. When the available passable width is higher than this threshold, the system will use the minimum weight value to reduce the impact of centerline deviation cost on the planning result.

[0074] In this embodiment, the piecewise function of the weight value can be expressed as: in, Here is the weight value, and width is the available passage width. Narrow threshold, For a wide threshold, The preset maximum weight, This is the preset minimum weight.

[0075] In this embodiment, by designing the weight value as a continuous piecewise function with the available passage width as the independent variable, precise control of the centerline deviation cost intensity is achieved. When the available passage width is lower than the narrow threshold, the weight value remains at its maximum value. At this time, the centerline deviation cost dominates the total cost, forcing the planned path to closely follow the centerline of the passable area, maximizing the safety of passage in narrow road sections. When the available passage width is higher than the wide threshold, the weight value drops to its minimum value, and the impact of the centerline deviation cost is correspondingly weakened. The algorithm can prioritize optimizing other performance indicators such as path length and smoothness. When the available passage width is in the intermediate transition range, the weight value decreases linearly with the increase of width, allowing the planning strategy to smoothly transition from "safety first" to "efficiency first". This continuous piecewise weight design avoids the jump in planning results caused by abrupt weight changes and ensures that the reference trajectory that best meets the safety requirements of the current passage conditions can be generated under different width scenarios, significantly improving the adaptability and reliability of the global path planning system.

[0076] Optionally, the weight value is an S-shaped function with the available passage width as the independent variable, and the weight value smoothly decreases from the maximum value to the minimum value as the available passage width increases.

[0077] In this embodiment, the S-shaped function is the Sigmoid function, which ensures that the weight value smoothly decreases from the maximum value to the minimum value according to the S-shaped curve as the available passage width increases, thus achieving continuous weight adjustment.

[0078] In this embodiment, by employing a sigmoid function as the weight calculation method, the weight value exhibits a smooth, non-linear decreasing characteristic as the available passage width increases, achieving continuous and natural adjustment of the centerline deviation cost intensity. This sigmoid function maintains a high weight value near the narrow threshold, ensuring sufficient centerline deviation penalty force when passage width is limited, forcing the planned path to closely adhere to the centerline of the passable area to guarantee passage safety. Near the wide threshold, the weight value gradually converges to a minimum, weakening the constraint on path centering and allowing the algorithm to prioritize optimizing other performance indicators such as path length and smoothness. In the intermediate transition region, the smoothness of the sigmoid function avoids trajectory jitter that may be caused by sudden weight changes, ensuring the stability of the reference trajectory generation. This sigmoid function-based weight adjustment mechanism enables the global planning algorithm to automatically adjust its optimization strategy according to changes in passage width, ensuring safe passage in narrow areas while fully leveraging the planning flexibility in wide areas, significantly improving the adaptability and robustness of the path planning system in different scenarios.

[0079] Optionally, obtain the reference trajectory of the target vehicle, including: Obtain the historical trajectory information recorded by the target vehicle during its historical driving process; The historical trajectory information is processed in reverse time sequence to obtain the reference trajectory.

[0080] In this embodiment of the application, the historical driving process refers to the driving process of the target vehicle over a period of time in the past.

[0081] In this embodiment of the application, the historical trajectory information is a sequence of trajectory points recorded by the target vehicle during its historical driving process, reflecting the vehicle's driving path.

[0082] In this embodiment of the application, the reverse time sequence processing refers to the processing operation of reversing the time sequence of historical trajectory information arranged in chronological order of historical driving time.

[0083] In this embodiment of the application, the reference trajectory refers to the sequence of trajectory points obtained by reverse time-series processing of historical trajectory information.

[0084] In this embodiment, a reference trajectory is generated by reverse time-series processing using historical trajectory information, enabling the target vehicle to plan based on the actual driving path during the reversing process, thereby improving the accuracy and reliability of the trajectory. At the same time, by introducing a passage width cost term into the local optimization algorithm with the goal of minimizing the objective function, and dynamically adjusting the cost intensity in the form of a power function of the remaining passage width, the passage width cost term increases when the remaining passage width decreases, thereby strengthening the obstacle avoidance constraint in narrow areas during the reversing process, ensuring that the vehicle maintains a safety margin when reversing, and improving driving safety and environmental adaptability.

[0085] Optionally, before performing trajectory optimization processing based on the vehicle state information, the reference trajectory, and the environmental perception data to obtain the trajectory optimization result of the reference trajectory, the method further includes: The reference trajectory is smoothed to obtain a smoothed reference trajectory; the smoothed reference trajectory is a trajectory with continuous curvature.

[0086] In this embodiment, the smoothing process refers to the optimization of the original reference trajectory using a curve fitting algorithm. Its purpose is to eliminate curvature abrupt changes in the original trajectory and generate a smooth path that satisfies vehicle kinematic constraints. The smoothing process can be achieved using common trajectory smoothing techniques such as polynomial interpolation, sliding window filtering, or spline curve fitting.

[0087] In this embodiment of the application, the smoothed reference trajectory is a trajectory curve with continuous curvature characteristics. This trajectory not only maintains the overall orientation characteristics of the original reference trajectory, but also ensures that the curvature change at any point on the trajectory is continuous and smooth.

[0088] In this embodiment, a trajectory with continuous curvature is obtained by smoothing the reference trajectory, effectively solving the problem of potential curvature abrupt changes in the original trajectory. Curvature continuity ensures smooth and consistent steering wheel operation, avoiding steering wheel vibration caused by discontinuous trajectory curvature and significantly improving ride comfort. Simultaneously, the smoothed reference trajectory provides input to the local optimization algorithm that better conforms to vehicle kinematics, resulting in smoother and more reasonable control parameters and improving the accuracy and stability of trajectory tracking. This preprocessing mechanism effectively reduces the complexity of subsequent optimization problems, contributing to improved performance and reliability of the entire trajectory planning system.

[0089] Optionally, the global planning algorithm is a hybrid A* algorithm.

[0090] In this embodiment, the Hybrid A* algorithm is referred to as the Hybrid A* Algorithm, which can include various variants of the algorithm. The Hybrid A* algorithm is a path planning method that combines the principles of traditional A* graph search with vehicle kinematic constraints, suitable for vehicle motion planning in a two-dimensional plane. Its principle lies in searching in a continuous state space, generating a smooth and feasible path that conforms to the vehicle's physical motion laws by considering the vehicle's kinematic model. The main steps of the Hybrid A* algorithm include: discretely generating subsequent possible state nodes based on the vehicle kinematic model; evaluating the estimated cost from the current node to the target node using a heuristic function; selecting the node with the minimum total cost for expansion, and performing collision detection during the expansion process; when the target region is found, backtracking to generate an initial path from the starting pose to the target pose by extracting the parent nodes of the expanded nodes; finally, the initial path is usually smoothed and optimized to further improve the path quality and the comfort of vehicle tracking.

[0091] In this embodiment, a hybrid A* algorithm is used to generate the reference trajectory through the global planning algorithm. It constructs a kinematically feasible path that meets the minimum turning radius constraint of the vehicle within the passable area. At the same time, it guides the path to be close to the centerline in narrow scenarios by using the centerline deviation cost term, providing high-quality initial trajectory input for local optimization and significantly reducing the risk of the vehicle scraping against obstacles.

[0092] Optionally, the local optimization algorithm is a model predictive control algorithm.

[0093] In this embodiment, the model predictive control algorithm is the Model Predictive Control (MPC) algorithm, which is an advanced control method based on the principle of rolling optimization. It predicts the state evolution sequence in the future finite time domain by establishing a vehicle kinematic model, solves the objective function optimization problem including the traffic width cost term in each control cycle, generates the optimal control parameter sequence at the current moment, and outputs the first term of the sequence as the actual control command to the actuator, thereby realizing closed-loop tracking control of the vehicle trajectory and real-time obstacle avoidance.

[0094] In this embodiment, the MPC algorithm is used as the local optimization algorithm. Leveraging its rolling optimization characteristics, it can solve for the objective function containing a power-law-type traffic width cost term based on real-time updated vehicle state, reference trajectory, and environmental perception data within each control cycle, generating the optimal control parameter sequence in real time. This algorithm incorporates the power-law-type traffic width cost term into the cost evaluation system within the prediction time domain. This allows the prediction model to anticipate the superlinear growth trend of the traffic width cost term when the remaining traffic width (the difference between available traffic width and vehicle width) shrinks in narrow road sections, thus proactively generating an obstacle avoidance trajectory. This combination of MPC and an objective function containing a sensitive traffic width cost term fully leverages the advantages of MPC in handling various constraints and dynamic optimization, while strengthening the precise control of the distance between the vehicle and obstacles through cost term design. Ultimately, this significantly improves the safety, real-time performance, and tracking accuracy of local trajectory planning, providing a reliable guarantee for the smooth and efficient driving of the target vehicle in various environments.

[0095] Exemplary device The apparatus embodiments of this application can be used to execute the method embodiments of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the method embodiments of this application.

[0096] Figure 2 The diagram shown is a block diagram of a vehicle trajectory planning device provided in one embodiment of this application. Figure 2 As shown, the device 200 includes: The acquisition module 210 is used to acquire vehicle status information, reference trajectory, and environmental perception data of the environment in which the target vehicle is located; the environmental perception data includes available passage width for characterizing the minimum distance between each location in the environment and its surrounding obstacles. The local optimization module 220 is used to perform trajectory optimization processing based on the vehicle status information, the reference trajectory, and the environmental perception data to obtain the trajectory optimization result of the reference trajectory. During the trajectory optimization process, the objective is to minimize the first traffic cost required to achieve the trajectory optimization result. The first traffic cost includes a traffic width cost term. The traffic width cost term is a power function with the remaining traffic width as the independent variable, and the traffic width cost term increases as the remaining traffic width decreases. The remaining traffic width is the difference between the available traffic width and the vehicle width of the target vehicle.

[0097] Optionally, the local optimization module 220 is used to solve the objective function to obtain the control parameters of the target vehicle with the goal of minimizing the value of the objective function of the local optimization algorithm; the objective function is used to represent the first passage cost required for the trajectory optimization result; the value of the objective function is calculated based on the vehicle state information, the reference trajectory, and the environmental perception data; the objective function includes a passage width cost term; when the remaining passage width exceeds the minimum safety margin, the passage width cost term is a power function with the remaining passage width as the independent variable.

[0098] Optionally, the acquisition module 210 is further configured to acquire the vehicle status information, target pose, and environmental perception data of the environment in which the target vehicle is located; wherein, the vehicle status information includes the initial pose of the target vehicle, and the environmental perception data includes information used to characterize the distribution of obstacles and the range of passable areas in the environment. The global planning module is used to perform trajectory planning based on the vehicle state information, the target pose, the environmental perception data, and the centerline, and generate a reference trajectory connecting the initial pose and the target pose. During the trajectory planning process, the objective is to minimize the second passage cost required for the reference trajectory. The second passage cost includes the centerline deviation cost. The centerline deviation cost is positively correlated with the deviation distance of the target vehicle from the centerline of the passable area at each node in the planned path.

[0099] Optionally, the global planning module is used to generate the reference trajectory based on a global planning algorithm with the goal of minimizing the second travel cost; wherein the second travel cost includes actual cost and heuristic cost, the actual cost is the cumulative travel cost of each node from the starting pose to the target pose, and the heuristic cost includes the centerline deviation cost.

[0100] Optionally, the centerline deviation cost is the product of the deviation distance and the weight value; the weight value is negatively correlated with the available passage width.

[0101] Optionally, the weight value is a continuous piecewise function with the available passage width as the independent variable; When the available passage width is less than a preset narrowing threshold, the weight value reaches its maximum value; When the available passage width is greater than a preset width threshold, the weight value reaches its minimum value; When the available passage width is between the narrow threshold and the wide threshold, the weight value decreases linearly from the maximum value to the minimum value as the available passage width increases.

[0102] Optionally, the weight value is an S-shaped function with the available passage width as the independent variable, and the weight value smoothly decreases from the maximum value to the minimum value as the available passage width increases.

[0103] Optionally, the acquisition module 210 is further configured to acquire historical trajectory information recorded by the target vehicle during its historical driving process; and to perform reverse time-series processing on the historical trajectory information to obtain the reference trajectory.

[0104] Optionally, the device 200 further includes: A smoothing module is used to smooth the reference trajectory to obtain a smoothed reference trajectory; the smoothed reference trajectory is a trajectory with continuous curvature.

[0105] Optionally, the global planning algorithm is a hybrid A* algorithm.

[0106] Optionally, the local optimization algorithm is a model predictive control algorithm.

[0107] Exemplary electronic devices Below, for reference Figure 3 This describes an electronic device according to embodiments of the present application. Figure 3 A block diagram of an electronic device according to an embodiment of this application is illustrated.

[0108] like Figure 3 As shown, the electronic device 300 includes one or more processors 310 and memory 320.

[0109] The processor 310 may be another form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 300 to perform desired functions.

[0110] Specifically, processor 310 can be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present invention. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. Processor 310 may also include a main processor, and may also include a baseband chip, a modem, etc.

[0111] The memory 320 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 310 may execute the program instructions to implement the vehicle trajectory planning methods of the various embodiments of this application described above and / or other desired functions. Various contents, such as category correspondences, may also be stored in the computer-readable storage medium.

[0112] In one example, the electronic device 300 may also include an input device 330 and an output device 340, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0113] In addition, the input device 330 can also be a device that receives user input data and information, such as a keyboard, mouse, camera, scanner, light pen, voice input device, touch screen, pedometer, or gravity sensor. The output device 340 can output various information to the outside. The output device 340 may include, for example, a display, speaker, printer, and communication network and its connected remote output devices.

[0114] Of course, for the sake of simplicity, Figure 3 Only some of the components of the electronic device 300 relevant to this application are shown in this illustration; components such as buses, input / output interfaces, etc., are omitted. In addition, the electronic device 300 may include any other suitable components depending on the specific application.

[0115] Exemplary vehicle In addition to the methods and devices described above, embodiments of this application may also include a vehicle, comprising a vehicle body and the electronic equipment.

[0116] Exemplary computer program products and computer-readable storage media In addition to the methods and devices described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the vehicle trajectory planning methods according to various embodiments of this application as described in the "Exemplary Methods" section of this specification.

[0117] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0118] Furthermore, embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps in the vehicle trajectory planning methods according to various embodiments of this application described in the "Exemplary Methods" section above.

[0119] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0120] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0121] For the foregoing method embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0122] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For apparatus embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0123] The steps in the methods of the various embodiments of this application can be adjusted, merged, or deleted in order according to actual needs, and the technical features described in each embodiment can be replaced or combined.

[0124] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0125] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0126] The modules or submodules described as separate components may or may not be physically separate. The components that constitute a module or submodule may or may not be physical modules or submodules; that is, they may be located in one place or distributed across multiple network modules or submodules. Some or all of the modules or submodules can be selected to achieve the purpose of this embodiment's solution, depending on actual needs.

[0127] Furthermore, the functional modules or sub-modules in the various embodiments of this application can be integrated into one processing module, or each module or sub-module can exist physically separately, or two or more modules or sub-modules can be integrated into one module. The integrated modules or sub-modules described above can be implemented in hardware or in the form of software functional modules or sub-modules.

[0128] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0129] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software unit executed by a processor, or a combination of both. The software unit can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0130] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0131] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A vehicle trajectory planning method, characterized in that, include: Acquire the vehicle status information and reference trajectory of the target vehicle, as well as the environmental perception data of the environment in which the target vehicle is located; The environmental perception data includes available passage widths used to characterize the minimum distance between each location in the environment and its surrounding obstacles; Based on the vehicle status information, the reference trajectory, and the environmental perception data, trajectory optimization processing is performed to obtain the trajectory optimization result of the reference trajectory; During the trajectory optimization process, the objective is to minimize the first passage cost required to achieve the trajectory optimization result; The first passage cost includes a passage width cost term; the passage width cost term is a power function with the remaining passage width as the independent variable, and the passage width cost term increases as the remaining passage width decreases; the remaining passage width is the difference between the available passage width and the vehicle width of the target vehicle.

2. The method according to claim 1, characterized in that, Based on the vehicle status information, the reference trajectory, and the environmental perception data, trajectory optimization processing is performed to obtain the trajectory optimization result of the reference trajectory, including: The objective function of the local optimization algorithm is minimized to obtain the control parameters of the target vehicle. The objective function represents the first passage cost required for the trajectory optimization result. The value of the objective function is calculated based on the vehicle state information, the reference trajectory, and the environmental perception data. The objective function includes a traffic width cost term; when the remaining traffic width exceeds the minimum safety margin, the traffic width cost term is a power function with the remaining traffic width as the independent variable.

3. The method according to claim 1, characterized in that, Obtain the reference trajectory of the target vehicle, including: The vehicle status information, target pose, and environmental perception data of the target vehicle's environment are acquired; wherein, the vehicle status information includes the target vehicle's initial pose, and the environmental perception data includes information characterizing the distribution of obstacles and the range of passable areas in the environment. Based on the environmental perception data, the centerline of the passable area in the environment where the target vehicle is located is determined; Based on the vehicle status information, the target pose, the environmental perception data, and the centerline, trajectory planning is performed to generate a reference trajectory connecting the starting pose and the target pose; during the trajectory planning process, the objective is to minimize the second passage cost required for the reference trajectory. The second passage cost includes a centerline deviation cost; the centerline deviation cost is positively correlated with the deviation distance of the target vehicle from the centerline of the passable area at each node in the planned path.

4. The method according to claim 3, characterized in that, The step of generating a reference trajectory connecting the initial pose and the target pose by performing trajectory planning based on the vehicle state information, the target pose, the environmental perception data, and the centerline includes: The reference trajectory is generated based on a global planning algorithm with the objective of minimizing the second passage cost. The second passage cost includes actual cost and heuristic cost. The actual cost is the cumulative travel cost of each node from the starting pose to the target pose, and the heuristic cost includes the centerline deviation cost.

5. The method according to claim 3, characterized in that, The centerline deviation cost is the product of the deviation distance and the weight value; the weight value is negatively correlated with the available passage width.

6. The method according to claim 5, characterized in that, The weight value is a continuous piecewise function with the available passage width as the independent variable; When the available passage width is less than a preset narrowing threshold, the weight value reaches its maximum value; When the available passage width is greater than a preset width threshold, the weight value reaches its minimum value; When the available passage width is between the narrow threshold and the wide threshold, the weight value decreases linearly from the maximum value to the minimum value as the available passage width increases.

7. The method according to claim 5, characterized in that, The weight value is an S-shaped function with the available passage width as the independent variable, and the weight value smoothly decreases from the maximum value to the minimum value as the available passage width increases.

8. The method according to claim 1, characterized in that, Obtain the reference trajectory of the target vehicle, including: Obtain the historical trajectory information recorded by the target vehicle during its historical driving process; The historical trajectory information is processed in reverse time sequence to obtain the reference trajectory.

9. The method according to claim 1, characterized in that, Before performing trajectory optimization processing based on the vehicle status information, the reference trajectory, and the environmental perception data to obtain the trajectory optimization result of the reference trajectory, the method further includes: The reference trajectory is smoothed to obtain a smoothed reference trajectory; the smoothed reference trajectory is a trajectory with continuous curvature.

10. A vehicle, characterized in that, Including electronic devices; The electronic device includes: a processor, and a memory for storing processor-executable instructions; The processor is configured to perform the method according to any one of claims 1 to 9.

Images