Automatic driving vehicle narrow road section U-turn path planning method and system

Abstract

The application provides a method and system for automatic driving vehicle U-turn path planning in narrow road sections, comprising: generating a multi-section hybrid A coarse path, the specific method comprising: extracting key pose feature points of the U-turn section based on environmental obstacles, lane lines and vehicle pose information, guiding a hybrid A algorithm to search for a feasible path with the feature points, triggering re-planning at a gear shifting and parking node, and outputting a U-turn coarse path meeting obstacle avoidance and kinematic constraints; optimizing the U-turn coarse path, the specific method comprising: taking the U-turn coarse path as a reference to construct an optimization boundary containing passable space constraints, lane line constraints and vehicle dynamics constraints, setting a multi-dimensional cost function of trajectory consistency, collision risk, smoothness and control smoothness, converting the path smoothing problem into an iterative linear quadratic regulator optimization proposition, and solving to obtain a smooth U-turn path with a control instruction. The application solves the problems of poor path compliance, discontinuous curvature and unfeasible dynamics of traditional algorithms.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving path planning technology, and in particular to a method and system for planning U-turn paths for autonomous vehicles in narrow road sections. Background Technology

[0002] With the large-scale deployment of autonomous driving technology, the ability to handle special conditions in complex urban traffic scenarios has become a core indicator for measuring the maturity of autonomous driving systems. U-turns in narrow road sections are a typical, high-frequency, and highly challenging scenario: when the road width is less than the vehicle's minimum turning diameter, the vehicle needs to complete the U-turn through multiple forward and reverse gear shifts, placing extremely high demands on obstacle avoidance safety, kinematic feasibility, and control smoothness in path planning. According to industry research, traditional autonomous driving systems face a collision risk rate as high as 34% when performing U-turns on roads less than 8 meters wide, with an average U-turn time of approximately 72 seconds, far exceeding the average of 28 seconds for human drivers. The limitations of planning algorithms directly restrict the usability of autonomous driving systems in complex road networks such as old urban areas and residential areas.

[0003] Current path planning schemes for U-turns on narrow roads mainly fall into two categories: the hybrid A* algorithm based on graph search and the fast random tree (RRT) algorithm based on sampling; among which:

[0004] The hybrid A* algorithm introduces a continuous state space on top of the traditional A* discrete grid search, connecting discrete nodes via Reeds-Shepp curves to handle vehicle steering constraints and gear shifting. Its execution flow is as follows: First, the map is discretized into a gridded cost map, defining constraints such as maximum front wheel angle, minimum turning radius, and allowed number of gear shifts, and setting the start and end poses; then, the core search loop is completed through node expansion and heuristic guidance, finally backtracking to the node sequence with the minimum cost to obtain the U-turn path. This scheme has the following inherent drawbacks:

[0005] Poor path curvature continuity: Discrete nodes are spliced ​​together by curves, and the path can only achieve positional continuity (C0 continuity), but cannot guarantee curvature continuity (C1 continuity). Curvature abrupt changes are likely to occur at the node connection points, causing the downstream control module to frequently turn the steering wheel to full and switch steering in the opposite direction, which not only reduces ride comfort, but also aggravates the wear of tires and steering mechanisms.

[0006] Ignoring vehicle dynamics constraints: The algorithm only considers kinematic geometric constraints and does not take into account dynamic characteristics such as tire slip ratio and steering system response delay. The planned path is theoretically collision-free, but in actual vehicle execution, it is prone to tracking lag and path deviation, which may lead to the risk of scraping the road boundary.

[0007] Uncontrollable waypoints and poor compliance: The algorithm only constrains the poses of the starting point and the end point, and there is no guidance mechanism for intermediate waypoints. It is easy to generate paths that violate road regulations and driving habits, such as crossing solid lines and taking shortcuts, which cannot meet the compliance requirements of mass-produced vehicles.

[0008] The RRT (Rapid Random Tree) algorithm explores feasible regions by randomly sampling in a high-dimensional state space. It constructs a random tree starting from the initial pose, successively expanding nodes until it approximates the target pose, and backtracks to obtain the turning path. It typically employs path pruning post-processing to optimize the path length. This approach has the following inherent drawbacks:

[0009] Poor path quality and significant non-optimality: Due to the inherent limitations of random sampling, the generated path usually contains a large number of redundant steerings and unnecessary gear shifts, resulting in poor smoothness and not conforming to the operating habits of human drivers.

[0010] Unstable computational efficiency: The algorithm only has probabilistic completeness. When the feasible region is narrow in a narrow space, the random sampling hit rate drops significantly, the algorithm's time consumption fluctuates greatly, and in extreme cases, it cannot output a feasible solution within the specified time, making it difficult to meet the real-time requirements of vehicle-mounted applications.

[0011] Lack of dynamic feasibility assurance: The search process does not incorporate vehicle dynamics constraints, the planned path is disconnected from the vehicle's execution capabilities, and problems such as the actuator exceeding the limit and the vehicle skidding are likely to occur during actual tracking.

[0012] In summary, existing technologies cannot simultaneously meet the four core requirements of path compliance, dynamic feasibility, control smoothness, and real-time computing efficiency in narrow road U-turn scenarios, becoming a key technological bottleneck restricting the implementation of autonomous driving in narrow scenarios. Summary of the Invention

[0013] This invention aims to solve at least one of the technical problems existing in the prior art, and proposes a method and system for planning U-turn paths for autonomous vehicles in narrow road sections.

[0014] In a first aspect, embodiments of the present invention provide a method for planning a U-turn path for an autonomous vehicle in a narrow road section, comprising:

[0015] The method for generating multi-segment hybrid A coarse path includes: extracting key pose feature points of the U-turn segment based on environmental obstacles, lane lines and vehicle pose information; using the feature points to guide the hybrid A algorithm to search for feasible paths; triggering replanning at the gear shift and stop node; and outputting a U-turn coarse path that satisfies obstacle avoidance and kinematic constraints.

[0016] The optimization of the coarse U-turn path includes the following methods: taking the coarse U-turn path as a reference, constructing an optimization boundary that includes constraints on passable space, lane lines, and vehicle dynamics; simultaneously setting a multi-dimensional cost function for trajectory consistency, collision risk, ride comfort, and control smoothness; transforming the path smoothing problem into an iterative linear quadratic regulator optimization problem; and solving for a smooth U-turn path with control commands.

[0017] Furthermore, output a coarse-grained turning path that satisfies obstacle avoidance and kinematic constraints. Specific methods include:

[0018] The road boundary, obstacle, and lane line information are discretized into a gridded cost map, and the state variables of the vehicle search node are initialized. The state variables include the vehicle position, heading angle, and driving speed.

[0019] By combining the solid and virtual attributes of lane lines with road sign and regulation information, the stop line, intersection center point, and start and end points of the turning arc in the U-turn area are identified as path guidance feature points.

[0020] Guided by feature points, nodes are expanded in the state space. Combining heuristic functions and vehicle minimum turning radius constraints, a collision-free initial path is searched.

[0021] When the vehicle reaches the shift point and is in a stopped state, a replanning is triggered with the current position as the new starting point and the center line of the target lane as the end point, and the subsequent path is updated.

[0022] By backtracking the optimal node sequence, a coarse path for turning around is generated, consisting of discrete pose points.

[0023] Furthermore, the current pose is used as the new starting point and the center line of the target lane is used as the endpoint to trigger replanning and update the subsequent path. The specific method includes: obtaining the deviation between the vehicle's heading angle and the target heading angle; when the deviation between the vehicle's heading angle and the target heading angle is greater than a preset threshold and the vehicle speed drops to zero, the replanning process and the vehicle gear shifting operation are performed simultaneously on the vehicle.

[0024] Furthermore, the optimization solution for the vehicular tunnel structure, where the passable space constraint extends along a coarse path, is performed only within the tunnel area, thereby reducing the solution space and improving computational efficiency.

[0025] Furthermore, the vehicle dynamics constraints include: maximum front wheel steering angle constraint, steering angle change rate constraint, longitudinal acceleration constraint, and tire slip ratio constraint.

[0026] Furthermore, the multi-dimensional cost function is specifically as follows:

[0027] The trajectory consistency cost specifically includes: measuring the lateral deviation between the optimized path and the coarse turning path;

[0028] The cost of collision risk specifically includes: measuring the distance between the optimized path and obstacles and road boundaries;

[0029] Smoothness costs specifically include: measuring the rate of change of path curvature and constraining lateral acceleration fluctuations;

[0030] Controlling the cost of smoothness includes: measuring the rate of change of steering wheel angle and acceleration, and constraining actuator impact.

[0031] Furthermore, the state variables for optimization by the iterative linear quadratic regulator are the vehicle pose and speed, and the control variables are the steering wheel angle and longitudinal acceleration. The solution process is based on iterative linearization of the vehicle's nonlinear dynamics model, and outputs the optimal path sequence and the corresponding control command sequence.

[0032] Furthermore, it also includes: sending a smooth U-turn path with accompanying control instructions to the vehicle chassis actuator, and executing the U-turn path planning through the vehicle chassis actuator.

[0033] Secondly, the present invention also discloses a path planning system for U-turns in narrow road sections for autonomous vehicles, comprising: a coarse path generation module and a path optimization module; wherein:

[0034] The coarse path generation module is used to extract key pose feature points of the U-turn section based on environmental obstacles, lane lines and vehicle pose information. The feature points guide the hybrid A algorithm to search for feasible paths. Replanning is triggered at the gear shift and stop node, and the coarse path of the U-turn that meets obstacle avoidance and kinematic constraints is output.

[0035] The path optimization module is used to construct an optimization boundary that includes traffic space constraints, lane line constraints and vehicle dynamics constraints, with the U-turn coarse path as a reference. At the same time, it sets a multi-dimensional cost function for trajectory consistency, collision risk, ride comfort and control smoothness, transforming the path smoothing problem into an iterative linear quadratic regulator optimization problem, and solving it to obtain a smooth U-turn path with control commands.

[0036] Thirdly, the present invention also discloses an electronic device, comprising:

[0037] One or more processors;

[0038] Memory, used to store one or more programs;

[0039] When the one or more programs are executed by the one or more processors, the one or more processors implement the method.

[0040] This invention provides a method and system for planning U-turn paths in narrow road sections for autonomous vehicles. The method includes: generating multiple segments of hybrid A-row coarse paths; specifically, extracting key pose feature points of the U-turn segment based on environmental obstacles, lane lines, and vehicle pose information; using these feature points to guide the hybrid A-row algorithm in searching for feasible paths; triggering replanning at a gear shift and stop node; and outputting a coarse U-turn path that satisfies obstacle avoidance and kinematic constraints. The method further includes optimizing the coarse U-turn path; specifically, constructing an optimization boundary that includes passable space constraints, lane line constraints, and vehicle dynamics constraints, using the coarse U-turn path as a reference; and setting multi-dimensional cost functions for trajectory consistency, collision risk, ride comfort, and control smoothness, transforming the path smoothing problem into an iterative linear quadratic regulator optimization problem, and solving for a smooth U-turn path with accompanying control commands. This invention solves the problems of poor path compliance, curvature discontinuity, and dynamic infeasibility in traditional algorithms.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1. Strong path compliance and in line with driving habits; by guiding the search direction through key feature points in the U-turn area, the problem of "taking shortcuts and crossing solid lines" in traditional algorithms is avoided from the root. The planned path strictly complies with road regulations and human driving habits, meeting the compliance requirements for the mass production of vehicles.

[0043] 2. Integrated planning and control with high dynamic feasibility: The path planning and control optimization are integrated into the model, incorporating complete vehicle dynamics constraints and outputting an executable trajectory with control commands. This solves the problems of disconnect between traditional planning and control and inaccurate path tracking, and significantly reduces the collision risk of U-turns in narrow road sections.

[0044] 3. Smooth path and good ride comfort: By optimizing the continuous curvature of the path and constraining the rate of change of steering wheel angle, the problem of frequently turning the steering wheel to its full extent and turning the wheel while stationary is avoided. This not only improves ride comfort but also reduces wear on the tires and steering mechanism and extends the service life of components.

[0045] 4. Gear shift replanning ensures a smooth and continuous process; by utilizing the time interval between gear shifts while stopped, replanning is performed, hiding the calculation time within the operation interval. This not only corrects driving errors but also avoids the vibrations caused by route switching during driving, making the smoothness of the U-turn process close to that of a human driver.

[0046] 5. High computational efficiency, meeting the real-time requirements of vehicle-mounted systems; the coarse planning stage accelerates the search by guiding with feature points, and the fine optimization stage narrows the solution space by using driving tunnels. With dual optimization, the time for a single complete planning can be controlled within 100ms, which is far superior to traditional algorithms and meets the requirements for real-time operation in vehicles. Attached Figure Description

[0047] Figure 1A flowchart illustrating a method for planning a U-turn route for an autonomous vehicle in a narrow road section, provided by an embodiment of the present invention;

[0048] Figure 2 A structural block diagram of an autonomous vehicle U-turn path planning system in narrow road sections provided by an embodiment of the present invention;

[0049] Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0050] To enable those skilled in the art to better understand the technical solutions of the present invention, exemplary embodiments of the present invention are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0051] Where there is no conflict, the various embodiments of the present invention and the features thereof may be combined with each other.

[0052] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0053] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Terms such as “connected” or “linked” are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.

[0054] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant art and the invention, and will not be interpreted as having an idealized or overly formal meaning unless expressly so defined herein.

[0055] In the technical solution of this invention, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information all comply with relevant laws and regulations and do not violate public order and good morals. The use of user data in this technical solution follows relevant national laws and regulations (e.g., the "Information Security Technology - Personal Information Security Specification"). For example: appropriate measures are taken for personal information access control; restrictions are imposed on the display of personal information; the purpose of using personal information does not exceed the scope of direct or reasonable association; and explicit identity targeting is eliminated when using personal information to avoid precisely locating a specific individual.

[0056] To address at least one of the technical problems existing in the aforementioned related technologies, the present invention provides a method and system for planning U-turn paths for autonomous vehicles in narrow road sections.

[0057] This implementation discloses a method for planning U-turn paths for autonomous vehicles in narrow road sections, such as... Figure 1 ,include:

[0058] S100. Generate a multi-segment hybrid A coarse path. The specific method includes: extracting key pose feature points of the U-turn segment based on environmental obstacles, lane lines and vehicle pose information, using the feature points to guide the hybrid A algorithm to search for feasible paths, triggering replanning at the gear shifting and stopping node, and outputting a U-turn coarse path that meets obstacle avoidance and kinematic constraints.

[0059] In S100 of this embodiment, a coarse-grained turning path that satisfies obstacle avoidance and kinematic constraints is output. The specific method includes:

[0060] S101. Environment Modeling and Node Initialization: Discretize the road boundary, obstacle and lane line information into a gridded cost map, and initialize the state variables of the vehicle search node, which include the vehicle position, heading angle and driving speed.

[0061] Specifically, environmental information such as road boundaries, static obstacles, and lane line virtual and real attributes are converted into a gridded cost map, and the state variables of vehicle search nodes are defined, including at least the vehicle's centroid coordinates, heading angle, and driving speed; at the same time, the starting pose, the ending target pose, and the vehicle's kinematic parameters are initialized.

[0062] S102. Key feature point extraction of U-turn sections: Combining the virtual and real attributes of lane lines with road sign and regulation information, identify the stop line, intersection center point, and start and end points of the turning arc in the U-turn area as path guidance feature points;

[0063] Specifically, by combining lane markings and road regulations, key structural points in the U-turn area are automatically identified as search guides, including but not limited to: the location of the stop line at the intersection, the geometric center point of the intersection, the start and end points of the turning arc, and the centerline anchor point of the target lane.

[0064] The extraction dimensions of feature points can be dynamically adapted according to business scenarios: for announcement verification scenarios, they are extracted according to key technology dimensions; for traffic certification scenarios, they are extracted according to operation report types; and for environmental protection monitoring scenarios, they are extracted according to actual measurement types, thus enabling one algorithm to adapt to multiple business scenarios.

[0065] S103. Hybrid A* Core Search Loop: Guided by feature points, expand nodes in the state space, combine heuristic functions and vehicle minimum turning radius constraints to search for a collision-free initial path;

[0066] Specifically, using the extracted feature points as intermediate guide points, the hybrid A* algorithm is used to expand the nodes in the state space one by one; the search process incorporates constraints such as the vehicle's minimum turning radius and maximum number of gear shifts, and combines heuristic functions to accelerate the search. Reeds-Shepp curves are used to connect adjacent discrete nodes, supporting forward and reverse gear switching, and generating an initial path that satisfies geometric obstacle avoidance and kinematic constraints.

[0067] S104. Shift Point Replanning: When the vehicle reaches the shift point and is in a stopped state, replanning is triggered with the current position as the new starting point and the center line of the target lane as the end point to update the subsequent path;

[0068] In this embodiment, replanning is triggered with the current pose as the new starting point and the center line of the target lane as the ending point to update the subsequent path. The specific method includes: obtaining the deviation between the vehicle's heading angle and the target heading angle; when the deviation between the vehicle's heading angle and the target heading angle is greater than a preset threshold and the vehicle speed drops to zero, the replanning process and the vehicle's gear shifting operation are performed simultaneously on the vehicle.

[0069] Specifically, when the vehicle reaches the preset shift point and its speed drops to zero, path replanning is triggered: using the vehicle's current actual position as the new starting point and the center line of the target lane as the heading constraint, a hybrid A* search is re-executed to update the subsequent U-turn path. The replanning process is performed synchronously with the vehicle's gear shifting operation, utilizing the time interval between stopping and shifting to complete the calculation, avoiding driving jitter caused by path switching during driving, and ensuring the smoothness and continuity of the entire U-turn process.

[0070] S105. Output Coarse-grained Turnaround Path: Backtrack the optimal node sequence to generate a coarse-grained turnaround path composed of discrete pose points. Specifically, backtrack the optimal node sequence and output a coarse-grained turnaround path composed of discrete pose points. This path meets basic obstacle avoidance and compliance requirements, but it has the characteristics of large curvature fluctuations and does not consider dynamic constraints, serving as a baseline reference line for subsequent optimization.

[0071] This S100 step outputs a preliminary path based on global environmental information that satisfies obstacle avoidance, kinematics, and compliance constraints, providing a high-quality reference for subsequent refined optimization.

[0072] S200. Optimize the coarse U-turn path. The specific method includes: taking the coarse U-turn path as a reference, constructing an optimization boundary that includes passable space constraints, lane line constraints and vehicle dynamics constraints, and setting a multi-dimensional cost function for trajectory consistency, collision risk, ride comfort and control smoothness, transforming the path smoothing problem into an iterative linear quadratic regulator optimization problem, and solving it to obtain a smooth U-turn path with control commands.

[0073] In this implementation S200, a multi-dimensional constraint boundary is first constructed, specifically including:

[0074] Passable space constraints: A strip-shaped driving tunnel is constructed along the U-turn coarse path. The tunnel width is set in combination with the vehicle width and the safety redundancy distance. All optimization solutions are performed within the tunnel area, which greatly reduces the solution space and improves computational efficiency.

[0075] Lane line constraints: Using solid road lines as hard constraint boundaries, ensuring that the optimized path does not cross the solid lines and complies with road traffic regulations.

[0076] Vehicle dynamics constraints: Incorporating dynamic constraints such as the vehicle's maximum front wheel steering angle, rate of change of steering angle, longitudinal acceleration, and tire slip ratio to ensure that the planning results can be accurately executed by the vehicle chassis.

[0077] Then, the multi-objective cost function is set, specifically including:

[0078] Trajectory consistency cost: measures the lateral deviation between the optimized path and the original coarse path, ensuring that the optimization does not deviate from the feasible region;

[0079] Collision risk cost: Measure the distance between waypoints and obstacles or road boundaries. The closer the distance, the higher the cost, forcing the path to maintain safety redundancy.

[0080] Ride comfort trade-off: Constraining path curvature and rate of change of curvature reduces lateral acceleration fluctuations and improves ride comfort;

[0081] Controlling the cost of smoothness: Constraining the rate of change of steering wheel angle and longitudinal acceleration, reducing actuator impact, and reducing tire and steering mechanism wear.

[0082] Finally, ILQR optimization is performed, transforming the path smoothing and control optimization problem into an iterative linear quadratic regulator (ILQR) optimization problem. Specific methods include:

[0083] State variables are defined as: vehicle center of mass coordinates, heading angle, and speed;

[0084] The control variables are defined as: steering wheel angle and longitudinal acceleration;

[0085] Based on the vehicle nonlinear dynamics model, iterative linearization is performed at each reference point, and the optimal control sequence and corresponding state sequence are solved by reverse recursion.

[0086] After iterative convergence, the final smooth U-turn path is output, along with the steering wheel angle and acceleration control commands corresponding to each path point.

[0087] Specifically, its cost function is:

[0088]

[0089] in:

[0090] Solving this problem yields the optimized U-turn path. The specific method is as follows:

[0091]

[0092] This design achieves integrated planning and control: the planning output is no longer a simple geometric path, but an executable trajectory with control instructions. Downstream control modules can directly track and execute it, effectively reducing tracking errors and avoiding safety risks caused by the disconnect between planning and control.

[0093] In some embodiments, a method for planning a U-turn path for an autonomous vehicle in a narrow road section further includes: S300. Sending a smooth U-turn path with control instructions to the vehicle chassis actuator, and performing the U-turn path planning through the vehicle chassis actuator.

[0094] This embodiment provides a method for planning U-turn paths in narrow road sections for autonomous vehicles, including: generating multiple segments of hybrid A-row coarse paths. Specifically, this method involves: extracting key pose feature points of the U-turn segment based on environmental obstacles, lane lines, and vehicle pose information; using these feature points to guide the hybrid A-row algorithm in searching for feasible paths; triggering replanning at a gear shift and stop node; and outputting a coarse U-turn path that satisfies obstacle avoidance and kinematic constraints. The method further involves optimizing the coarse U-turn path. Specifically, this method involves: using the coarse U-turn path as a reference, constructing an optimization boundary that includes passable space constraints, lane line constraints, and vehicle dynamics constraints; and setting multi-dimensional cost functions for trajectory consistency, collision risk, ride comfort, and control smoothness. This transforms the path smoothing problem into an iterative linear quadratic regulator optimization problem, solving for a smooth U-turn path with accompanying control commands. This invention solves the problems of poor path compliance, curvature discontinuity, and dynamic infeasibility in traditional algorithms.

[0095] Based on the same inventive concept, embodiments of the present invention also provide a path planning system for autonomous vehicles to make U-turns in narrow road sections, such as... Figure 2 It includes: a coarse path generation module and a path optimization module; wherein:

[0096] The coarse path generation module is used to extract key pose feature points of the U-turn section based on environmental obstacles, lane lines and vehicle pose information. The feature points guide the hybrid A algorithm to search for feasible paths. Replanning is triggered at the gear shift and stop node, and the coarse path of the U-turn that meets obstacle avoidance and kinematic constraints is output.

[0097] The path optimization module is used to construct an optimization boundary that includes traffic space constraints, lane line constraints and vehicle dynamics constraints, with the U-turn coarse path as a reference. At the same time, it sets a multi-dimensional cost function for trajectory consistency, collision risk, ride comfort and control smoothness, transforming the path smoothing problem into an iterative linear quadratic regulator optimization problem, and solving it to obtain a smooth U-turn path with control commands.

[0098] In this embodiment, an autonomous vehicle narrow road U-turn path planning system further includes an interactive output module, which is used to send a smooth U-turn path with control commands to the vehicle chassis actuator, and execute the U-turn path planning through the vehicle chassis actuator.

[0099] The specific working methods of the coarse path generation module, path optimization module, and interactive output module have been described in detail in the above-mentioned method for planning U-turn paths in narrow road sections for autonomous vehicles, and will not be repeated here.

[0100] Based on the same inventive concept, embodiments of the present invention also provide an electronic device. Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device including: one or more processors 101, a memory 102, and one or more I / O interfaces 103. The memory 102 stores one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement any of the methods described in the above embodiments; the one or more I / O interfaces 103 are connected between the processor and the memory, configured to enable information interaction between the processor and the memory.

[0101] The processor 101 is a device with data processing capabilities, including but not limited to a central processing unit (CPU); the memory 102 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and flash memory (FLASH); the I / O interface (read / write interface) 103 is connected between the processor 101 and the memory 102, and can realize information interaction between the processor 101 and the memory 102, including but not limited to a data bus (Bus).

[0102] In some embodiments, the processor 101, memory 102, and I / O interface 103 are interconnected via bus 104, and thus connected to other components of the computing device.

[0103] In some embodiments, the one or more processors 101 include a field-programmable gate array.

[0104] This invention also provides a computer-readable medium. The computer-readable medium stores a computer program, which, when executed by a processor, implements the steps of any of the methods described in the above embodiments. The computer-readable storage medium may be a volatile or non-volatile computer-readable storage medium.

[0105] This invention also provides a computer program product, including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code, wherein when the computer-readable code is run in the processor of an electronic device, the processor in the electronic device executes the above-described method.

[0106] Those skilled in the art will understand that all or some of the steps, systems, and apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software can be distributed on a computer-readable storage medium, which may include computer storage media (or non-transitory media) and communication media (or transient media).

[0107] As is known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable program instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), static random access memory (SRAM), flash memory or other memory technologies, portable compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, it is known to those skilled in the art that communication media typically contain computer-readable program instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0108] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0109] The computer program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the computer-readable program instructions. This electronic circuitry can execute the computer-readable program instructions to implement various aspects of the invention.

[0110] The computer program product described herein can be implemented specifically through hardware, software, or a combination thereof. In one alternative embodiment, the computer program product is specifically embodied in a computer storage medium; in another alternative embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0111] Various aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0112] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0113] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0114] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction, which contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0115] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A method for planning a U-turn path for an autonomous vehicle in a narrow road section, characterized in that, include: The method for generating multi-segment hybrid A coarse path includes: extracting key pose feature points of the U-turn segment based on environmental obstacles, lane lines and vehicle pose information; using the feature points to guide the hybrid A algorithm to search for feasible paths; triggering replanning at the gear shift and stop node; and outputting a U-turn coarse path that satisfies obstacle avoidance and kinematic constraints. The optimization of the coarse U-turn path includes the following methods: taking the coarse U-turn path as a reference, constructing an optimization boundary that includes constraints on passable space, lane lines, and vehicle dynamics; simultaneously setting a multi-dimensional cost function for trajectory consistency, collision risk, ride comfort, and control smoothness; transforming the path smoothing problem into an iterative linear quadratic regulator optimization problem; and solving for a smooth U-turn path with control commands.

2. The method according to claim 1, characterized in that, Output a coarse-grained U-turn path that satisfies obstacle avoidance and kinematic constraints. Specific methods include: The road boundary, obstacle, and lane line information are discretized into a gridded cost map, and the state variables of the vehicle search node are initialized. The state variables include the vehicle position, heading angle, and driving speed. By combining the solid and virtual attributes of lane lines with road sign and regulation information, the stop line, intersection center point, and start and end points of the turning arc in the U-turn area are identified as path guidance feature points. Guided by feature points, nodes are expanded in the state space. Combining heuristic functions and vehicle minimum turning radius constraints, a collision-free initial path is searched. When the vehicle reaches the shift point and is in a stopped state, a replanning is triggered with the current position as the new starting point and the center line of the target lane as the end point, and the subsequent path is updated. By backtracking the optimal node sequence, a coarse path for turning around is generated, consisting of discrete pose points.

3. The method according to claim 2, characterized in that, The current position is used as the new starting point and the center line of the target lane is used as the end point to trigger replanning and update the subsequent path. The specific method includes: obtaining the deviation between the vehicle's heading angle and the target heading angle. When the deviation between the vehicle's heading angle and the target heading angle is greater than a preset threshold and the vehicle speed drops to zero, the replanning process and vehicle gear shifting operation are performed simultaneously on the vehicle.

4. The method according to claim 1, characterized in that, The optimization solution for the vehicular tunnel structure with the passable space constraint extending along the coarse path is performed only within the tunnel area, which is used to reduce the solution space and improve computational efficiency.

5. The method according to claim 1, characterized in that, The vehicle dynamics constraints include: maximum front wheel steering angle constraint, steering angle change rate constraint, longitudinal acceleration constraint, and tire slip rate constraint.

6. The method according to claim 1, characterized in that, The multi-dimensional cost function is specifically as follows: The trajectory consistency cost specifically includes: measuring the lateral deviation between the optimized path and the coarse turning path; The cost of collision risk specifically includes: measuring the distance between the optimized path and obstacles and road boundaries; Smoothness costs specifically include: measuring the rate of change of path curvature and constraining lateral acceleration fluctuations; Controlling the cost of smoothness includes: measuring the rate of change of steering wheel angle and acceleration, and constraining actuator impact.

7. The method according to claim 1, characterized in that, The state variables for optimization by the iterative linear quadratic regulator are vehicle pose and speed, and the control variables are steering wheel angle and longitudinal acceleration. The solution process is based on iterative linearization of the vehicle's nonlinear dynamics model, and outputs the optimal path sequence and the corresponding control command sequence.

8. The method according to claim 1, characterized in that, Also includes: The smooth U-turn path with accompanying control commands is sent to the vehicle chassis actuator, which then executes the U-turn path planning.

9. A path planning system for U-turns in narrow road sections for autonomous vehicles, employing any one of the methods described in claims 1-8, characterized in that, include: The module consists of a coarse path generation module and a path optimization module; where: The coarse path generation module is used to extract key pose feature points of the U-turn section based on environmental obstacles, lane lines and vehicle pose information. The feature points guide the hybrid A algorithm to search for feasible paths. Replanning is triggered at the gear shift and stop node, and the coarse path of the U-turn that meets obstacle avoidance and kinematic constraints is output. The path optimization module is used to construct an optimization boundary that includes traffic space constraints, lane line constraints and vehicle dynamics constraints, with the U-turn coarse path as a reference. At the same time, it sets a multi-dimensional cost function for trajectory consistency, collision risk, ride comfort and control smoothness, transforming the path smoothing problem into an iterative linear quadratic regulator optimization problem, and solving it to obtain a smooth U-turn path with control commands.

10. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 8.

Images