A cross-zone transfer path planning method, device and medium for a hinged road roller under a limited working condition

Abstract

This invention relates to a method, device, and medium for cross-regional transfer path planning of an articulated roller under constrained working conditions. The method includes the following steps: obtaining the state of the latest node in the current search tree as the current vehicle state vector of the articulated roller, and constructing a kinematic model; calculating the current forward reachable sector during the path search process based on the current vehicle state vector, and generating a reference trajectory under constrained working conditions based on the forward reachable sector; obtaining a predicted trajectory based on the reference trajectory and the kinematic model using an improved CL-RRT trajectory closed-loop prediction mechanism, and adding the predicted trajectory as a new node to the search tree, with the articulated roller moving according to the predicted trajectory; repeating the above steps until the articulated roller enters the target work area, completing the cross-regional transfer path planning process. Compared with the prior art, this invention has the advantages of significantly improving planning efficiency and maneuver success rate in confined spaces.

Description

Technical Field

[0001] This invention relates to the field of traffic engineering, and in particular to a method, equipment, and medium for cross-regional transfer path planning of articulated road rollers under constrained working conditions. Background Technology

[0002] With the continuous improvement of mechanization in road construction, articulated rollers, with their unique articulated steering structure, small turning radius, and high maneuverability, have become key equipment in high-grade highway construction. Currently, research on automated roller operation mainly focuses on compaction quality control and basic motion control, while research on cross-work zone transfer path planning under constrained working conditions is relatively limited.

[0003] However, articulated rollers face complex mobility challenges in actual construction operations. Vehicles often need to move between areas in narrow, obstacle-filled confined spaces, requiring precise forward-reverse maneuvers. On one hand, the confined space imposes extremely stringent geometric constraints on the path, forcing planning algorithms to find feasible solutions within a very small free space. On the other hand, articulated vehicles suffer from nonholonomic constraints and maximum articulation angle limitations, making it difficult to directly execute simple geometric paths, thus increasing the difficulty of planning and control.

[0004] In terms of path planning technology, existing solutions mainly have the following limitations: First, the kinematic modeling is too simplified. Many general algorithms ignore the maximum articulation angle limit and nonholonomic constraints unique to articulated vehicles, resulting in large tracking errors in actual execution of the generated paths, or even failure to execute due to exceeding physical limits. Second, there is a lack of a systematic reversing decision mechanism for narrow working conditions. Existing methods often treat forward and reverse as independent processes, making it difficult to intelligently determine the optimal reversing switching point, leading to operational redundancy or planning failure. Third, planning and control are separated. Traditional sampling-based planning methods are often detached from the underlying tracking controller. Although the generated trajectories are geometrically connected, they may be infeasible at the dynamic level, and often ignore the fine-grained requirements for end attitude (such as heading alignment and articulation angle zeroing), causing the vehicle to be unable to directly connect to the next work process.

[0005] In summary, the core contradiction of existing technologies lies in the mismatch between the high mobility and high-precision transfer requirements of articulated rollers under constrained working conditions and the existing planning methods' "inaccurate constraint modeling, simplistic decision-making mechanisms, and separation of planning and control." Therefore, there is an urgent need to develop a path planning method to enhance the autonomous transfer capabilities of rollers in complex construction environments. Summary of the Invention

[0006] The purpose of this invention is to provide a method, equipment, and medium for cross-regional transfer path planning of articulated rollers under restricted working conditions, which significantly improves the feasibility and safety of the path in narrow working conditions.

[0007] The objective of this invention can be achieved through the following technical solutions: A method for cross-regional transfer path planning of an articulated road roller under constrained working conditions, comprising the following steps: Obtain the state of the latest node in the current search tree, use it as the current vehicle state vector of the articulated roller, and construct a kinematic model; Based on the current vehicle state vector, the current forward reachable sector is calculated during the path search process, and a reference trajectory under constrained conditions is generated based on the forward reachable sector. Based on the reference trajectory and kinematic model, a predicted trajectory is obtained by using an improved CL-RRT trajectory closed-loop prediction mechanism, and the predicted trajectory is added as a new node to the search tree. The articulated roller moves according to the predicted trajectory. Repeat the above steps until the articulated roller enters the target work area, completing the cross-regional transfer route planning process.

[0008] Furthermore, the process of constructing the kinematic model includes: The current vehicle state vector is set as follows: ,in The position of the front wheel center in the global coordinate system. The heading angle of the front wheels. The angle between the longitudinal axes of the front and rear wheels, i.e., the hinge angle; A kinematic model relating the current vehicle state vector to the control input, with the front wheel center as the reference point, is established as follows: , In the formula, The center of the front wheel is located at the lower edge of the global coordinate system. shaft and The velocity component in the axial direction, The rate of change of the vehicle's heading angle. The rate of change of the hinge angle. The linear velocity of the front wheel. For hinge angular velocity, This is the distance from the center of the front wheel to the hinge point.

[0009] Furthermore, the calculation steps for the forward reachable sector include: The maximum allowable articulation angle of the articulated roller is set as follows: ; Based on the current vehicle state vector and the maximum permissible articulation angle Calculate the geometric boundaries that the articulated roller can reach in the current direction of motion; The sector-shaped region formed by the geometric boundary and the longitudinal axis of the articulated roller is defined as the forward reachable sector, wherein the angle of the forward reachable sector is limited by the maximum trajectory curvature constraint.

[0010] Furthermore, the step of generating the reference trajectory includes: Determine the sampling target point Whether it is located within the forward reachable sector, If so, then generate a connection between the current node and the sampling target point. The forward motion trajectory segment serves as a reference trajectory for the forward motion. The current node is the latest node in the search tree, representing the current state node of the articulated roller during the path planning process. If not, then select a point on the boundary of the forward reachable sector as the reversing switching point. Generate a connection between the current node and the reversing switch point. The forward trajectory segment is generated, and the connection between the reversing switching points is established. With the sampling target point The reversing trajectory segments form a forward-reversing composite trajectory, which serves as a reference trajectory for the composite motion.

[0011] Furthermore, the reversing switching point The selection steps include: Calculate the distance between each point on the boundary of the forward reachable sector and the sampling target point. The distance is calculated, and a point on the boundary that minimizes this distance is selected as the reversing switching point. .

[0012] Furthermore, the step of generating the predicted trajectory includes: Based on the reference trajectory, the tracking controller calculates the control input at each moment; The control input is substituted into the kinematic model for forward integration simulation, and the constraints to be satisfied are detected in real time during the simulation to obtain a predicted trajectory that satisfies all constraints. The end of the predicted trajectory is a safe stopping state. The constraints include the following: (1) Hinge angle Not exceeding the maximum allowable range , The maximum allowable hinge angle; (2) The trajectory curvature shall not exceed the minimum turning radius limit of the articulated roller; (3) The trajectory point will not collide with the obstacle.

[0013] Furthermore, the process of the articulated roller moving according to the predicted trajectory also includes the following steps: If multiple predicted trajectories are generated during the planning process, the predicted trajectories are used as candidate paths, and the comprehensive cost value of each candidate path is calculated. The comprehensive cost value is obtained by weighting the length of the candidate path, the number of turning operations, and the minimum distance to the obstacle. Set end attitude constraints, including: the deviation between the heading angle and the rolling direction of the target work area is less than a preset threshold, and the articulation angle converges to near zero; Determine whether a candidate path that satisfies the end attitude constraint can be found within the planning time. If so, the articulated roller moves according to the candidate path. If not, select the candidate path with the lowest comprehensive cost and a safe stop state at the end as the alternative path. After the articulated roller travels along the alternative path to the safe stop point, it triggers an emergency stop or replans.

[0014] Furthermore, it also includes adjusting the predicted trajectory of the articulated roller when it enters the target work area, specifically including the following steps: For the predicted trajectory of entering the target work area, the relative target state and initial articulation angle deviation of the articulated roller are calculated. The relative target state includes the relative position between the current position and the target work area, the deviation between the current heading angle and the rolling direction of the target work area, and the deviation between the current articulation angle and the target zero. The relative target state and initial articulation angle deviation are input into the generator of the trained generative adversarial network (GAN) model, which outputs a control sequence that satisfies the end-effector attitude constraints of the target work area. This control sequence includes the steering angle and duration of multiple control segments. During training, the discriminator simulates an articulated roller executing the control sequence generated by the generator and determines whether the end-effector state after execution satisfies the end-effector attitude constraints of the target work area. The control sequence is derived using the kinematic model to obtain a guiding path, and the movement is performed according to the guiding path to complete the trimming process.

[0015] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the cross-regional transfer path planning method for an articulated road roller under constrained working conditions as described above.

[0016] The present invention also provides a storage medium storing a program thereon, which, when executed, implements a cross-regional transfer path planning method for an articulated road roller under constrained working conditions as described above.

[0017] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention replaces the traditional fixed-direction search and calculates the reachable area in real time according to the current state, so that the generated reference trajectory is more in line with the motion characteristics of the articulated roller. Especially in the case of limited turning radius and narrow space, it can effectively avoid the problem of trajectory infeasibility or interference with the environment. Furthermore, it introduces an improved CL-RRT trajectory closed-loop prediction mechanism to perform closed-loop prediction of the reference trajectory, ensuring that the planned predicted trajectory is not only theoretically reachable, but can also be accurately executed by the vehicle's kinematic model, which greatly improves the reliability of the trajectory and the actual following accuracy, thereby significantly improving the feasibility and safety of the path in narrow working conditions.

[0018] (2) This invention addresses the unique kinematic characteristics of articulated rollers by integrating the unique kinematic constraints of articulated rollers (such as articulation angle limits and maximum trajectory curvature constraints) throughout the entire process from model construction to trajectory prediction. This ensures that the planned path is consistent with the actual motion capability of the vehicle and conforms to the actual maneuverability of the vehicle. It avoids the generation of infeasible paths from the source and also avoids the execution error caused by simplifying the vehicle into a point model.

[0019] (3) This invention innovatively proposes a reversing point decision method based on forward reachability analysis, realizing the organic integration of reversing operation and trajectory generation, which significantly improves the planning efficiency and maneuver success rate in narrow spaces.

[0020] (4) The present invention adopts the improved CL-RRT algorithm, which deeply couples trajectory generation with underlying tracking control. The closed-loop prediction mechanism ensures the physical executability of the planned trajectory and solves the problem of separation between planning and control.

[0021] (5) The present invention introduces a generative adversarial network model to refine the end trajectory, which meets the special operational requirements of the road roller for heading alignment and articulation angle to zero when entering a new target work area, and improves the level of construction automation.

[0022] (6) The safety stop mechanism integrated in this invention enhances the robustness of the algorithm, provides reliable emergency protection in complex environments, and avoids the vehicle from falling into an uncontrollable state.

[0023] (7) This invention allows the algorithm to perform real-time replanning through incremental search and iterative optimization, which can cope with temporary obstacles or changes in working conditions that occur dynamically on the construction site without global recalculation and has a fast response speed. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is the forward reachable sector of the present invention; Figure 3 This is an experimental scenario for an embodiment of the present invention; Figure 4 The experimental test results are for embodiments of the present invention. Detailed Implementation

[0025] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0026] Example 1 This embodiment provides a method for cross-regional transfer path planning of articulated road rollers under constrained working conditions, such as... Figure 1 As shown, the method includes the following steps: S1: Construct the kinematic model of a bidirectional drive articulated roller.

[0027] This embodiment acquires real-time environmental information of the current construction site (such as obstacle locations, road boundaries, temporary facilities, etc.) and the initial state of the bidirectional articulated roller. And initialize the search tree.

[0028] In this embodiment, the double-drum articulated roller consists of a front wheel, a rear wheel, and a central articulation device. The kinetic model is specifically a double-drum articulated model, and the model parameters include the distance from the center of the front wheel to the articulation point. and the distance from the center of the rear wheel to the hinge point The kinematic model satisfies nonholonomic constraints, meaning it assumes the steel wheel does not sideslip and its lateral velocity is zero. The current vehicle state vector (represented by the state of the latest node in the current search tree) is defined as follows: ,in The position of the front wheel center in the global coordinate system. The heading angle of the front wheels. The angle between the longitudinal axes of the front and rear wheels; the control input during vehicle movement is the linear velocity of the front wheels. and hinge angular velocity By controlling the hinge angle The change in the steering motion is achieved through the change in the control input. Therefore, a kinematic model relating the state vector with the front wheel center as a reference point and the control input is established as follows: , This model not only describes the motion patterns of vehicles, but also provides a basis for subsequent accessibility analysis.

[0029] S2: Reversing point decision and compound trajectory generation based on reachability analysis.

[0030] For path planning under constrained working conditions, this embodiment proposes a reversing point selection mechanism based on reachability analysis. During the path search process, the current vehicle state vector and the maximum allowable articulation angle are first considered. Calculate the geometric boundaries that the vehicle can reach in its current direction of motion. The sector formed by this boundary and the vehicle's longitudinal axis is taken as the current forward reachable sector. For example... Figure 2 As shown, this sector represents the geometric boundary that the vehicle can reach while maintaining forward motion mode, and its sector angle is limited by the maximum trajectory curvature.

[0031] When sampling target point When the node is located outside the sector, a connection between the current node and the sampling target point is directly generated. The forward motion trajectory segment.

[0032] When sampling target point When the vehicle is located outside this fan-shaped area, it indicates that forward movement alone is insufficient to reach the target, necessitating the introduction of reverse maneuvers. In this case, the reverse switching point is determined using the following method. : Calculate the distance between each point on the reachable sector boundary and the target point, and select the boundary point with the smallest distance as the reversing switching point.

[0033] Subsequently, a connection is generated between the current node and the reversing switch point. The forward trajectory segment, and the connection to the reversing switch point. With the sampling target point The reversing trajectory segments form a forward-reversing composite trajectory.

[0034] It can be seen that the generation of the forward-reverse composite trajectory is divided into two stages: the first stage is when the vehicle moves in forward mode to... In the second phase, Switching to reverse mode, the center of the rear wheels is used as the new tracking reference point to control the rear wheels to track the planned path until the target area is reached. This method effectively avoids the problem of excessive articulation angle caused by forced steering in narrow spaces, ensuring the physical feasibility of the path.

[0035] S3: Trajectory loop prediction and feasibility screening based on improved CL-RRT.

[0036] In the path generation phase, this embodiment employs an improved CL-RRT algorithm, introducing a closed-loop prediction mechanism. In each node expansion of the random search tree, this mechanism does not directly sample randomly in the control input space, but instead generates a reference trajectory as an intermediary. Specifically: Using the forward trajectory segment and forward-reverse composite trajectory found during the path search process as the reference trajectory, a preset low-level tracking controller is invoked. The control input at each moment is calculated based on the reference trajectory and then input into the kinematic model for forward integration simulation, resulting in a predicted trajectory containing a series of continuous states. Simultaneously, the following constraints are monitored in real-time during the simulation: Hinge Does it exceed the maximum allowed range? ; Does the trajectory curvature exceed the vehicle's minimum turning radius limit? Does the trajectory point collide with an obstacle? If all the above constraints are met, the predicted trajectory is added as a new node to the random tree; otherwise, the trajectory segment is discarded.

[0037] In addition, this step introduces trajectory cost evaluation and a safe stopping mechanism to ensure that the end of each trajectory segment is in a "safe stopping state," enhancing the robustness of the system. When there are multiple predicted trajectories that satisfy all constraints (collision constraints, kinematic constraints, and safe stopping constraints), each predicted trajectory in the search tree is taken as a candidate path, and a comprehensive cost value is calculated, where the comprehensive cost value is a weighted value of the candidate path length, the number of turning operations, and the minimum distance to the obstacle. During trajectory simulation, ensure that the end of each trajectory segment is a "safe stopping state", meaning that the vehicle has the ability to stop safely at that point; If a candidate path that satisfies the end attitude constraints can be found within the planning time, the articulated roller will move along the candidate path. If not, the candidate path with the lowest comprehensive cost and a safe stop at the end is selected as the alternative path. After the articulated roller travels along the alternative path to the safe stop point, an emergency stop or replanning is triggered. The end attitude constraints include: the deviation between the heading angle and the rolling direction of the target work area is less than a preset threshold, and the articulation angle converges to near zero.

[0038] S4: End-of-line attitude optimization of road rollers based on generative adversarial networks (GANs).

[0039] Repeat steps S1-S3 until the articulated roller enters the target work area, and obtain the predicted trajectory of the articulated roller entering the target work area. This step optimizes the attitude of the predicted trajectory and uses the trained generative adversarial network (GAN) model to refine the predicted trajectory.

[0040] The GAN model consists of a generator network and a discriminator network. The generator takes the vehicle's relative target state and initial articulation angle deviation as input and outputs a discretized steering control sequence, which includes the steering angle and duration of multiple control segments. The discriminator simulates the execution of the control sequence output by the generator and determines whether the final state after execution meets the final attitude constraints of the target work area (the deviation between the heading angle and the rolling direction of the target work area is less than a preset threshold, and the articulation angle converges to near zero).

[0041] Using a trained generator, the predicted trajectory planned by CL-RRT is optimized to generate a guide path that allows the roller to smoothly enter the target work area with a zero articulation angle. Specific steps include: For the predicted trajectory of entering the target work area, calculate the vehicle's relative target state, including the relative position of the current position and the target point of the work area, the deviation of the current heading angle from the rolling direction of the work area, the deviation of the current articulation angle from the target (zero), and extract the initial articulation angle deviation. The relative target state and initial hinge angle deviation are input into the trained generator, which outputs a control sequence that satisfies the end attitude constraints of the target work area. The kinematic model is used to deduce the control sequence to obtain the guide path, so that the movement is carried out according to the guide path to complete the trimming process.

[0042] In summary, the above method calculates the forward reachable sector based on the vehicle's current state vector at the current node. Based on this sector, it determines whether to proceed directly forward or introduce a backtracking operation to solve trajectory planning in narrow working conditions. Furthermore, based on an improved CL-RRT trajectory closed-loop prediction mechanism, trajectory generation is deeply coupled with the underlying tracking control. This closed-loop prediction mechanism ensures the physical executability of the planned trajectory, solving the problem of separating planning and control. Finally, a generative adversarial network model is used to refine the end-point trajectory, ensuring that the articulated roller meets the special operational requirements of heading alignment and zeroing the articulation angle when entering a new work area. This achieves cross-regional transfer path planning for the articulated roller, improving the level of construction automation. To verify the practical effectiveness of the above method, a real-vehicle experiment was conducted on a section of highway. The experimental platform used a modified double-drum roller (BOMAG) with the following parameters: wheelbase 3300mm, distance from wheel center to hinge point 1.6m, and maximum hinge angle ±40°. Figure 3 As shown, the experimental scenario involves a transfer task between two independent work areas separated by obstacles. The experimental results show that the vehicle can track the planned path well, with an average tracking error of 14.44 cm. The tracking history curve is shown below. Figure 4 As shown, the vehicle smoothly switches motion modes at the reversing switching point and finally enters the target work area with a heading angle close to 180° and an articulation angle close to zero, verifying the feasibility and robustness of the method of the present invention under actual complex working conditions.

[0043] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0044] Example 2 This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a cross-regional transfer path planning method for an articulated road roller under constrained working conditions as described in Embodiment 1 above.

[0045] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0046] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will 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 program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0047] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0048] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0049] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0050] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for cross-regional transfer path planning of an articulated road roller under constrained working conditions, characterized in that, Includes the following steps: Obtain the state of the latest node in the current search tree, use it as the current vehicle state vector of the articulated roller, and construct a kinematic model; Based on the current vehicle state vector, the current forward reachable sector is calculated during the path search process, and a reference trajectory under constrained conditions is generated based on the forward reachable sector. Based on the reference trajectory and kinematic model, a predicted trajectory is obtained by using an improved CL-RRT trajectory closed-loop prediction mechanism, and the predicted trajectory is added as a new node to the search tree. The articulated roller moves according to the predicted trajectory. Repeat the above steps until the articulated roller enters the target work area, completing the cross-regional transfer route planning process.

2. The method for cross-regional transfer path planning of an articulated road roller under constrained working conditions according to claim 1, characterized in that, The process of constructing the kinematic model includes: The current vehicle state vector is set as follows: ,in The position of the front wheel center in the global coordinate system. The heading angle of the front wheel. The angle between the longitudinal axes of the front and rear wheels, i.e., the hinge angle; A kinematic model relating the current vehicle state vector to the control input, with the front wheel center as the reference point, is established as follows: ， In the formula, The front wheel center at the bottom edge of the global coordinate system shaft and The velocity component in the axial direction, The rate of change of the vehicle's heading angle. The rate of change of the hinge angle. The linear velocity of the front wheel. For hinge angular velocity, This is the distance from the center of the front wheel to the hinge point.

3. The method for cross-regional transfer path planning of an articulated road roller under constrained working conditions according to claim 1, characterized in that, The calculation steps for the forward reachable sector include: The maximum allowable articulation angle of the articulated roller is set as follows: ; Based on the current vehicle state vector and the maximum permissible articulation angle Calculate the geometric boundaries that the articulated roller can reach in the current direction of motion; The sector-shaped region formed by the geometric boundary and the longitudinal axis of the articulated roller is defined as the forward reachable sector, wherein the angle of the forward reachable sector is limited by the maximum trajectory curvature constraint.

4. The method for cross-regional transfer path planning of an articulated road roller under constrained working conditions according to claim 1, characterized in that, The steps for generating the reference trajectory include: Determine the sampling target point Whether it is located within the forward reachable sector, If so, then generate a connection between the current node and the sampling target point. The forward motion trajectory segment serves as a reference trajectory for the forward motion, wherein the current node is the latest node in the search tree, representing the current state node of the articulated roller during the path planning process; If not, then select a point on the boundary of the forward reachable sector as the reversing switching point. Generate a connection between the current node and the reversing switch point. The forward trajectory segment is generated, and the connection between the reversing switching points is established. With the sampling target point The reversing trajectory segments form a forward-reversing composite trajectory, which serves as a reference trajectory for the composite motion.

5. The method for cross-regional transfer path planning of an articulated road roller under constrained working conditions according to claim 4, characterized in that, The reversing switching point The selection steps include: Calculate the distance between each point on the boundary of the forward reachable sector and the sampling target point. The distance is calculated, and a point on the boundary that minimizes this distance is selected as the reversing switching point. .

6. The method for cross-regional transfer path planning of an articulated road roller under constrained working conditions according to claim 1, characterized in that, The steps for generating the predicted trajectory include: Based on the reference trajectory, the tracking controller calculates the control input at each moment; The control input is substituted into the kinematic model for forward integration simulation, and the constraints to be satisfied are detected in real time during the simulation to obtain a predicted trajectory that satisfies all constraints. The end of the predicted trajectory is a safe stopping state. The constraints include the following: (1) Hinge angle Not exceeding the maximum allowable range , The maximum allowable hinge angle; (2) The trajectory curvature shall not exceed the minimum turning radius limit of the articulated roller; (3) The trajectory point will not collide with the obstacle.

7. The method for cross-regional transfer path planning of an articulated road roller under constrained working conditions according to claim 6, characterized in that, The articulated roller also includes the following steps during its movement along the predicted trajectory: If multiple predicted trajectories are generated during the planning process, the predicted trajectories are used as candidate paths, and the comprehensive cost value of each candidate path is calculated. The comprehensive cost value is obtained by weighting the length of the candidate path, the number of turning operations, and the minimum distance to the obstacle. Set end attitude constraints, including: the deviation between the heading angle and the rolling direction of the target work area is less than a preset threshold, and the articulation angle converges to near zero; Determine whether a candidate path that satisfies the end attitude constraint can be found within the planning time. If so, the articulated roller moves according to the candidate path. If not, select the candidate path with the lowest comprehensive cost and a safe stop state at the end as the alternative path. After the articulated roller travels along the alternative path to the safe stop point, it triggers an emergency stop or replans.

8. The method for cross-regional transfer path planning of an articulated road roller under constrained working conditions according to claim 1, characterized in that, It also includes adjusting the predicted trajectory of the articulated roller when it enters the target work area, specifically including the following steps: For the predicted trajectory of entering the target work area, the relative target state and initial articulation angle deviation of the articulated roller are calculated. The relative target state includes the relative position between the current position and the target work area, the deviation between the current heading angle and the rolling direction of the target work area, and the deviation between the current articulation angle and the target zero. The relative target state and initial articulation angle deviation are input into the generator of the trained generative adversarial network (GAN) model, which outputs a control sequence that satisfies the end-effector attitude constraints of the target work area. This control sequence includes the steering angle and duration of multiple control segments. During training, the discriminator simulates an articulated roller executing the control sequence generated by the generator and determines whether the end-effector state after execution satisfies the end-effector attitude constraints of the target work area. The control sequence is derived using the kinematic model to obtain a guiding path, and the movement is performed according to the guiding path to complete the trimming process.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements a cross-regional transfer path planning method for an articulated road roller under constrained working conditions as described in any one of claims 1 to 8.

10. A storage medium having a program stored thereon, characterized in that, When the program is executed, it implements a cross-regional transfer path planning method for an articulated road roller under restricted working conditions as described in any one of claims 1-8.

Images