Task-oriented safe autonomous stringing vehicle path planning hierarchical cooperative control system

Abstract

The present application belongs to the technical field of autonomous overhead line car path planning, and discloses a task safety-oriented autonomous overhead line car path planning hierarchical cooperative control system, which solves the problems that the traditional algorithm is easy to fall into local optimum and dead loop and ignores cable tension safety during path planning. The system comprises a task decision layer, an algorithm planning layer and a terminal execution layer which are in turn hierarchically cooperative. The task decision layer takes a task safety finite state machine as the core. The algorithm planning layer fuses a JPS-PLUS global path planning algorithm, a local path optimization algorithm coupled with a cable tension model and a real-time obstacle identification module, and sets a cable tension safety threshold to complete a safety path calculation task. The terminal execution layer is responsible for trajectory obstacle avoidance and path finding according to the algorithm planning layer, safe wiring and real-time feedback. The present application provides effective protection for the safety and continuity of autonomous overhead line car task operation.

Description

Technical Field

[0001] This invention belongs to the field of autonomous overhead line vehicle path planning technology, and particularly relates to a hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety. Background Technology

[0002] Currently, an increasing number of intelligent unmanned equipment uses wireless communication as their primary means of communication. However, in emergency communication scenarios in the field, the environment is complex with obstacles, communication infrastructure is easily damaged, and wireless communication methods are at risk of failure. Therefore, intelligent equipment capable of quickly laying and repairing cables has become an important component of field communication support. Autonomous cable laying vehicles (hereinafter referred to as autonomous cable laying vehicles) are one of the core pieces of equipment capable of completing this task, and their path planning is crucial for safe operation. Therefore, it is essential to research safe and reliable path planning algorithms for autonomous cable laying vehicles.

[0003] Currently, path planning for mobile intelligent agents often employs hybrid path planning algorithms that combine global and local planning. Dijkstra's algorithm is one of the most commonly used global path planning algorithms. However, it operates on two-dimensional discrete graphs and only considers the actual cost from the starting point to a node, lacking consideration of the actual situation, thus having significant limitations. An extension of Dijkstra's algorithm is the A* algorithm, which introduces a heuristic function to estimate the cost to the target point, making the path search directional. However, in complex obstacle environments, the design difficulty of its heuristic function increases, limiting its application. Further extensions of A* lead to the Jump Point Search (JPS) algorithm. This algorithm skips a large number of redundant nodes, searching only for key "jump points," thus its computational speed is significantly faster than A*. It is suitable for scenarios with high real-time requirements and is widely used in AGV grid navigation, global static path planning for mobile intelligent agents, and pathfinding on large maps.

[0004] Local path planning (DWA) dynamically avoids sudden obstacles during the movement of a mobile agent based on real-time surrounding information, and performs real-time correction and local optimization of the global reference trajectory. It is a key technology to ensure the agent's safe movement, autonomous obstacle avoidance, and smooth path finding. DWA, based on robot kinematic constraints, samples and filters in the velocity space, exhibiting good performance in terms of real-time performance, computational complexity, and robustness. It can balance obstacle avoidance safety, trajectory smoothness, and heading accuracy, and is widely used in scenarios such as local real-time obstacle avoidance, trajectory following, and autonomous navigation for mobile agents. However, traditional DWA algorithms are prone to getting trapped in local optima. To address this problem, many scholars have conducted extensive optimization research on improving evaluation functions, optimizing constraints, and integrating other algorithms.

[0005] In terms of the integration of global path planning and local path planning algorithms, most studies focus on optimizing path smoothness and obstacle avoidance efficiency, and have certain limitations in their attention to the continuity of mobile agent tasks and the safety of algorithm operation. For example, there is a lack of extrication compensation mechanisms for local optimal deadlock and iterative dead loop. In addition, the movement of autonomous wiring vehicles is strictly constrained by the tension of wiring cables. If cable tension safety is ignored, it can easily lead to work interruption, cable damage, and communication failure. Summary of the Invention

[0006] One objective of this invention is to provide a hierarchical collaborative control system for path planning of autonomous wire-laying vehicles oriented towards mission safety, which effectively solves the problems that traditional algorithms are prone to getting stuck in local optima and iterative dead loops when planning paths for autonomous wire-laying vehicles, and that neglect cable tension safety.

[0007] To address the aforementioned technical problems, the present invention adopts the following technical solution: A hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety comprises a task decision layer, an algorithm planning layer, and a terminal execution layer that work in a hierarchical manner. The task decision layer uses a mission safety finite state machine as its core, standardizes the division and switching logic of work states, realizes abnormal working condition identification and self-rescue, and ensures the continuity of mission operations. The algorithm planning layer integrates the JPS-PLUS global path planning algorithm, the DWA local path optimization algorithm coupled with the cable tension model, and the YOLO-based real-time obstacle recognition module, and sets a cable tension safety threshold to complete the safe path calculation task. The terminal execution layer is responsible for obstacle avoidance and path finding according to the trajectory issued by the algorithm planning layer, safe wiring, and real-time feedback.

[0008] Furthermore, the cost function of the JPS-PLUS global path planning algorithm :

[0009] ;

[0010] ;

[0011] ;

[0012] ;

[0013] In the formula, Indicates the distance from the starting point to the current node. The actual cost, Indicates the current node Heuristic cost estimation to the destination This represents the straight-line distance from the current node to the destination. This represents the straight-line distance from the starting point to the ending point. This represents the density of obstacles in the rectangular region formed by the current node and the target node. This indicates the number of obstacles within the rectangular area formed by the current node and the target node. This indicates the size of the rectangular region formed by the current node and the target node. The vector consisting of the starting point and the target node. The vector formed by the starting point and the current node cross product, Representing vectors with vector The angle between them.

[0014] Furthermore, compared to the traditional JPS algorithm, the JPS-PLUS global path planning algorithm adds detection of diagonal obstacles and pruning of redundant nodes in the path. The method for pruning redundant nodes is as follows: determine whether there are obstacle grids in the grid traversed by the line connecting two nodes; if not, delete all redundant nodes between the two nodes. Assuming the node array storing the path is... Define a direct line-of-sight detection function between two nodes. and other node detection functions between two nodes :

[0015] ;

[0016] ;

[0017] In the formula, For the node with the smaller array index, For nodes with larger array indices, , .

[0018] Furthermore, the steps for direct line-of-sight detection between two nodes and detection of other nodes are as follows: S1, Initialization, let... , S2, Connecting Node , Generate line segments S3, Calling the function If the return value is 1, it means there are other nodes between the two nodes, proceed to step S4; if the return value is 0, it means there are no other nodes between the two nodes, obtain the pruned path and update it. Proceed to step S6; S4, Call the function If the return value is 1, it means that the two nodes are not directly visible to each other, and the line segment... After passing through the obstacle, proceed to step S6; if the return value is 0, it means that a line segment can be directly connected between the two nodes. If no obstacle is traversed, delete the path node array. , All nodes between, get Let the number of nodes , Proceed to step S5; S5, let , :like To obtain the path after pruning End the detection; otherwise, proceed to step S3; S6, let , Proceed to step S3.

[0019] Furthermore, the DWA local path optimization algorithm for the coupled cable tension model, compared to the traditional DWA algorithm, modifies the motion constraints, i.e., adds tension constraints after coupling the cable tension:

[0020] ;

[0021] In the formula, This indicates the acceleration of the moving line of the autonomous wire-laying vehicle; Indicates the maximum allowable tension of the cable; This indicates the cable tension at the connection point between the cable and the autonomous cable-laying vehicle. This indicates the density of the cable's own material; Represents gravitational acceleration; This indicates the length of the suspended section of the cable from the connection point on the autonomous cable-laying vehicle to the ground contact section; Indicates the quality of the independently developed wire-laying vehicle; This indicates the vertical height of the suspended section, which is the height of the connection point between the cable and the autonomous cable-laying vehicle from the ground. Indicates the current total cable length; Indicates the minimum cable length.

[0022] Modify dynamic windows:

[0023] ;

[0024] In the formula, This indicates the improved dynamic window. Indicates linear velocity. Indicates angular velocity, This indicates the linear velocity of the autonomous wire-laying vehicle. This represents the maximum linear acceleration after considering cable tension constraints. This indicates the angular velocity of the autonomous wire-laying vehicle. This represents the maximum angular acceleration after considering cable tension constraints. Indicates the sampling time interval.

[0025] Update speed sampling range:

[0026] For each sampling rate pair Calculate the corresponding and And only if the condition is satisfied At that time, the calculated and Only then was it retained for use.

[0027] ;

[0028] ;

[0029] In the formula, Indicates the dynamic tension of the cable. Indicates the static tension of the cable. This indicates the acceleration of cable laying.

[0030] Furthermore, compared to the traditional DWA algorithm, the DWA local path optimization algorithm with coupled cable tension model modifies the trajectory calculation formula to a piecewise uniform velocity model under cable tension constraints.

[0031] ;

[0032] ;

[0033] In the formula, express The linear velocity of motion at any given moment. express of Axial component, express of Axial component, express The heading angle at any moment, express angular velocity at any given moment express The linear velocity of motion at any given moment. express of Axial component, express of Axial component, express The heading angle at any given moment.

[0034] Update the comprehensive evaluation function:

[0035] ;

[0036] ;

[0037] In the formula, This represents the comprehensive evaluation function. This represents the heading angle evaluation function. This represents the obstacle distance evaluation function. This represents the current velocity evaluation function. This indicates the function for tracing the global path evaluation. This represents the tension cost evaluation function. , , , , These represent the weight values ​​of the corresponding evaluation functions. Represents the normalization factor. This represents the tension weighting coefficient. This indicates the cable tension safety threshold.

[0038] Furthermore, the task safety finite state machine's operation process is divided into five states: S0, S1, S2, S3, and S4. S0 represents standby, S1 represents global path planning, S2 represents local path optimization, S3 represents the autonomous wiring vehicle's obstacle avoidance and pathfinding, and S4 represents self-rescue. When the autonomous wiring vehicle is in S0, if it detects "receiving work instructions," it executes "loading map and wiring task information" and transitions to S1. When the autonomous wiring vehicle is in S1, if it detects "no reachable path," it executes "warning of abnormal task termination and requesting obstacle clearing" and transitions to S0. If it detects "a reachable path," it executes "using the global path as an evaluation sub-function parameter to guide local path optimization" and transitions to S2. When the autonomous wiring vehicle is in S2, if it detects "working conditions and path calculations are normal," it executes "setting cable tension safety threshold." The system guides the autonomous wiring vehicle to find its way, and then jumps to S3. If it detects that the path calculation is stuck in a local optimum, it executes "mark the location information of the local optimum on the map" and jumps to S4. When the autonomous wiring vehicle is in S3, if it detects that "the destination has been reached", it executes "mark the task as completed normally" and jumps to S0. If it detects that "the destination has not been reached and no new obstacle has been encountered", it executes "continue to find its way" and jumps to S3. If it detects that "the destination has not been reached and a new obstacle has been encountered", it executes "mark the location of the new obstacle on the map and start the 'escape strategy'" and jumps to S4. When the autonomous wiring vehicle is in S4, if it detects that "the escape has been successful", it executes "continue to find its way" and jumps to S3. If it detects that "the escape has failed", it executes "mark the location of the new obstacle on the map and start the 'recovery mechanism'" and jumps to S1.

[0039] Furthermore, when the autonomous wiring vehicle is in S4 self-rescue mode, it activates the "escape strategy" and comprehensively evaluates the function. :

[0040] ;

[0041] in, This represents the obstacle distance evaluation function in a self-rescue state. This represents the velocity evaluation subfunction under self-rescue conditions. , , These represent the weight values ​​of the corresponding evaluation functions under the self-rescue state.

[0042] In a self-rescue state, the closer you get to the minimum safe distance... Obstacle distance evaluation subfunction The higher the value, the faster the speed; the speed evaluation subfunction. The higher the value, the more the autonomous wire-laying vehicle moves along the edge of the obstacle.

[0043] Activate the "escape strategy" and use Simultaneously, the original comprehensive evaluation function for each sampling point is recorded. The resulting evaluation value When it appears ,illustrate The new predicted trajectory obtained at time t is already better than The time trajectory has a higher comprehensive evaluation value. The autonomous wire-laying vehicle has successfully escaped the predicament. The mission safety finite state machine has switched from state S4 to state S3, and the self-rescue has ended.

[0044] Furthermore, when the autonomous wiring vehicle gets stuck in a local minimum, it activates an "escape strategy" and records the current coordinates as follows: Simultaneously calculate the previous guiding path node. With the next bootstrap node distance When the actual distance traveled exceeds twice the speed again And again If the "escape strategy" fails, the "recovery mechanism" is activated, the current operating condition information is recorded, and the system enters the S1 state.

[0045] Compared with existing technologies, the beneficial technical effects of this invention are as follows: This invention provides effective assurance for the safety and continuity of autonomous wire-laying vehicle operations through a hierarchical collaborative control framework and a hybrid path planning algorithm. The hierarchical collaborative control framework can effectively avoid the risk of path planning failure, ensuring the safety of autonomous wire-laying vehicle operations; the hybrid path planning algorithm exhibits good performance in terms of task completion rate, cable tension safety, and dynamic obstacle avoidance success rate. This invention can provide a technical reference for path planning and task safety control of intelligent support equipment for field emergency communication. Attached Figure Description

[0046] Figure 1 This is a diagram of the hierarchical collaborative control system architecture for the autonomous overhead line vehicle path planning of the present invention.

[0047] Figure 2 It is the state transition diagram of a task-safe finite state machine.

[0048] Figure 3 This is the result of running the traditional JPS algorithm on a small-sized map with high obstacle density.

[0049] Figure 4 This is the result of running the JPS-PLUS algorithm on a small-sized map with high obstacle density.

[0050] Figure 5 This is the result of running the traditional JPS algorithm on a medium-sized map with high obstacle density.

[0051] Figure 6 This is the result of running the JPS-PLUS algorithm on a medium-sized map with high obstacle density.

[0052] Figure 7 This is the result of running the traditional JPS algorithm on a large-size map with high obstacle density.

[0053] Figure 8 This is the result of running the JPS-PLUS algorithm on a large-size map with high obstacle density.

[0054] Figure 9 It is a trajectory map of the autonomous cable-laying vehicle in a small-sized map with low obstacle density.

[0055] Figure 10 This is a graph showing the changes in the linear velocity and actual laying speed of the autonomous wire-laying vehicle in a small-sized map with low obstacle density.

[0056] Figure 11 It is a trajectory map of the autonomous cable-laying vehicle in a small-sized map with high obstacle density.

[0057] Figure 12 This is a graph showing the changes in the linear velocity and actual line laying speed of the autonomous line-laying vehicle in a small-sized map with high obstacle density.

[0058] Figure 13 It is a simulation diagram of the autonomous power line vehicle's trajectory under conditions of information updates in a small-sized, low-obstacle-density map.

[0059] Figure 14 This is a graph showing the changes in the linear velocity and actual laying speed of the autonomous wire-laying vehicle during the simulation process, under conditions where information is updated on a small-sized map with low obstacle density.

[0060] Figure 15It is a trajectory diagram of the autonomous power line vehicle during the simulation process, under the condition that information is updated on a medium-sized map with low obstacle density.

[0061] Figure 16 This is a graph showing the changes in the linear velocity and actual laying speed of the autonomous wire-laying vehicle during the simulation process, under conditions where information is updated on a medium-sized map with low obstacle density.

[0062] Figure 17 This is a trajectory map of the autonomous wire-laying vehicle during an "escape strategy" test experiment on a small-sized map with high obstacle density.

[0063] Figure 18 This is a graph showing the changes in the linear velocity and actual line laying speed of the autonomous wire-laying vehicle during an "escape strategy" test experiment on a small-sized map with high obstacle density.

[0064] Figure 19 This is a trajectory diagram of the autonomous wire-laying vehicle during a "recovery mechanism" test experiment conducted on a small-sized map with high obstacle density.

[0065] Figure 20 This is a graph showing the changes in the linear velocity and actual line laying speed of the autonomous line-laying vehicle during a "recovery mechanism" test experiment on a small-sized map with high obstacle density. Detailed Implementation

[0066] Example 1: This example provides a hierarchical collaborative control system for path planning of an autonomous overhead line vehicle, oriented towards mission safety. It includes a hierarchically collaborative task decision-making layer, an algorithm planning layer, and a terminal execution layer. The hierarchical collaborative control structure of "task decision-algorithm planning-terminal execution" provided in this example has the advantages of layered decision-making, step-by-step coordination, and closed-loop feedback. It can decouple complex planning and control problems into three levels: decision-making, planning, and execution, meeting the control requirements of autonomous overhead line vehicles (characterized by high mission continuity requirements, complex operational constraints, dynamic and variable operating conditions, and susceptibility to local optima).

[0067] like Figure 1 As shown in Table 1, the task decision layer (top layer) uses a task safety finite state machine as its core, standardizing the division and switching logic of work states, realizing abnormal working condition identification and self-rescue, and ensuring the continuity of task operation. The algorithm planning layer (middle layer) integrates the JPS-PLUS global path planning algorithm, the DWA local path optimization algorithm coupled with the cable tension model, and the YOLO-based real-time obstacle recognition module, and sets the cable tension safety threshold to complete the safe path calculation task. The terminal execution layer (bottom layer) is responsible for obstacle avoidance and pathfinding according to the trajectory issued by the algorithm planning layer, safe wiring, and real-time feedback of working conditions, motion status, and cable status.

[0068] Table 1. Task Definitions for Each Level of the Hierarchical Cooperative Control System for Path Planning of Autonomous Cable Car

[0069]

[0070] 1. Design of a task-safe finite state machine for the task decision layer.

[0071] The path planning and control of the autonomous wire-laying vehicle should prioritize mission safety, achieving closed-loop management of the entire operation process. To this end, a mission safety finite state machine is designed, dividing the operation process into five states. The meanings, transition rules, triggering events, and actions of each state are shown in Table 2. Figure 2 As shown, the task safety finite state machine's operation flow is divided into five states: S0, S1, S2, S3, and S4. S0 represents standby, S1 represents global path planning, S2 represents local path optimization, S3 represents the autonomous wiring vehicle's obstacle avoidance and pathfinding, and S4 represents self-rescue. ① When the autonomous wiring vehicle is in S0, if it detects "receiving a work instruction," it executes "loading the map, task information, etc." and transitions to S1. ② When the autonomous wiring vehicle is in S1, if it detects "no reachable path," it executes "warning of abnormal task termination and requesting obstacle clearing" and transitions to S0; if it detects "a reachable path," it executes "using the global path as an evaluation sub-function parameter to guide local path optimization" and transitions to S2. ③ When the autonomous wiring vehicle is in S2, if it detects that "the working condition and path calculation are normal", it executes "set the cable tension safety threshold to guide the autonomous wiring vehicle to find its way" and jumps to S3; if it detects that "the path calculation is trapped in a local optimum", it executes "mark the location trapped in the local optimum on the map" and jumps to S4. ④ When the autonomous wiring vehicle is in S3, if it detects that "the destination has been reached", it executes "report that the task has ended normally" and jumps to S0; if it detects that "the destination has not been reached and no new obstacle has been encountered", it executes "continue to find its way" and jumps to S3; if it detects that "the destination has not been reached and a new obstacle has been encountered", it executes "mark the new obstacle on the map and activate the 'escape strategy'" and jumps to S4. ⑤ When the autonomous wiring vehicle is in S4, if it detects that "the escape has been successful", it executes "continue to find its way" and jumps to S3; if it detects that "the escape has failed", it executes "mark the location of the new obstacle on the map and activate the 'recovery mechanism'" and jumps to S1.

[0072] As the logical core of the mission decision-making layer, the mission safety finite state machine enables the autonomous wire-laying vehicle to be fully controllable throughout the entire operation process, identify faults, and perform self-rescue in case of anomalies, thus ensuring the continuous and uninterrupted wire-laying mission from the top-level mechanism.

[0073] Table 2. State Transition Table of the Task-Safe Finite State Machine

[0074]

[0075] 2. Hybrid Path Planning Algorithm Design for the Algorithm Planning Layer. This embodiment designs a hybrid path planning algorithm for the safe path problem, including: the JPS-PLUS global path planning algorithm, the DWA local path optimization algorithm coupled with the cable tension model, and a YOLO-based real-time obstacle recognition module. The real-time obstacle recognition module is based on the YOLOv11 model. This embodiment focuses on path planning and tension constraint optimization, and the principles of the YOLOv11 algorithm will not be elaborated further.

[0076] 2.1 JPS-PLUS Global Path Planning Algorithm. This embodiment improves upon three problems of the traditional JPS algorithm to form the JPS-PLUS algorithm.

[0077] (1) To address the problem that the traditional JPS algorithm will cross diagonally intersecting and adjacent obstacles when performing jump point expansion, a jump point expansion rule is added.

[0078] The traditional JPS algorithm's jump-point expansion mechanism does not support diagonal search (assuming this direction is the upper right). When obstacles exist simultaneously to the right and above the current position, the algorithm will directly skip the obstacles and check the next grid node in this direction. This is obviously unacceptable for real autonomous wiring vehicles. Therefore, a jump-point expansion rule is added to increase the detection of diagonal obstacles. The diagonal jump-point expansion rule is supplemented as follows: Before performing a diagonal search, a joint detection of the horizontal and vertical adjacent grids of the current node is added; when both vertical neighbors in the target diagonal direction are detected as obstacles, the current diagonal expansion is forcibly terminated, and further jumps in that direction are prohibited to avoid the path directly crossing the obstacle area.

[0079] (2) To address the problem that the heuristic function of the traditional JPS algorithm is too simple and has low versatility and practicality, the heuristic function is improved.

[0080] The cost function in this embodiment :

[0081] ;

[0082] ;

[0083] ;

[0084] ;

[0085] In the formula, Indicates the distance from the starting point to the current node. The actual cost; Indicates the current node Heuristic cost estimation to the destination; This represents the straight-line distance from the current node to the endpoint. Indicates the straight-line distance from the starting point to the ending point; This represents the density of obstacles in the rectangular region formed by the current node and the target node. This indicates the number of obstacles within the rectangular area formed by the current node and the target node; This indicates the size of the rectangular region formed by the current node and the target node; Representing vectors with vector The angle between them; Indicates the starting point and target node The vector formed From the starting point and the current node The vector formed The larger the cross product, the worse the current search direction and the higher the deviation from the target point; otherwise, the better the search direction and the lower the deviation from the target point.

[0086] When the algorithm initially searches and is close to the starting point, it aims to find a path to the target point as quickly as possible to accelerate progress. However, as the search position gets closer to the target, it assumes that the target is near and shifts its focus from rapid progress to finding a more refined and concise path to the target. This is due to the nodes... The value of is within the range [0,1]. According to the above expression, the shortest distance between two points is a straight line, so the straight-line distance from the starting point to the ending point is the expected shortest and best path. When When =0.5, weight values ​​and With equal weights, when the distance to the target point exceeds half the shortest distance, the goal is to accelerate the approach to the target; conversely, when the distance is less than half the shortest distance, the goal is to find a shorter and more precise search path. This is achieved by increasing obstacle density. As a weighting factor, it emphasizes that when the obstacle density is too high, it tends to find a shorter and better path, avoiding the possibility of finding a non-optimal path due to rushing towards the goal. This indicates that the higher the obstacle density, The larger the weight, the less impact the search direction has on the overall cost function. This embodiment improves the algorithm's efficiency by modifying the heuristic function, effectively reducing unnecessary node searches and significantly reducing the management of redundant nodes.

[0087] (3) To address the problem that the path obtained by the traditional JPS algorithm has too many inflection points, resulting in insufficient smoothness and excessive jaggedness, which affects the pathfinding efficiency of the autonomous wire-laying vehicle, redundant nodes in the path are pruned.

[0088] The basic idea of ​​path redundancy node pruning is to determine whether there are obstacle grids in the grid traversed by the line connecting two nodes. If not, all redundant nodes between the two nodes are deleted.

[0089] Assuming the node array of the storage path is Define a direct line-of-sight detection function between two nodes. and other node detection functions between two nodes :

[0090] ;

[0091] ;

[0092] In the formula, For the node with the smaller array index, For nodes with larger array indices, , .

[0093] The steps for direct line-of-sight detection between two nodes and detection of other nodes are as follows: S1, Initialization, let , S2, Connecting Node , Generate line segments S3, Calling the function If the return value is 1, it means there are other nodes between the two nodes, proceed to step S4; if the return value is 0, it means there are no other nodes between the two nodes, obtain the pruned path and update it. Proceed to step S6; S4, Call the function If the return value is 1, it means that the two nodes are not directly visible to each other, and the line segment... After passing through the obstacle, proceed to step S6; if the return value is 0, it means that a line segment can be directly connected between the two nodes. If no obstacle is traversed, delete the path node array. , All nodes between, get Let the number of nodes , Proceed to step S5; S5, let , :like To obtain the path after pruning End the detection; otherwise, proceed to step S3; S6, let , Proceed to step S3.

[0094] 2.2 Local Path Optimization Algorithm for DWA Coupled with Cable Tension Model. Traditional DWA algorithms only consider the basic kinematic constraints and obstacle avoidance of the mobile robot. When applied to the motion planning of an autonomous wire-laying vehicle, cable tension should be considered as an additional constraint. Therefore, this embodiment makes the following improvements to the traditional DWA algorithm.

[0095] (1) Modify the motion constraint, that is, add tension constraint after coupling cable tension. The main physical parameters related to cable tension are defined as shown in Table 3.

[0096] Table 3 Definitions of key physical parameters related to cable tension

[0097]

[0098] The kinematic constraints of the autonomous wire-laying vehicle considered in the traditional DWA algorithm are:

[0099] ;

[0100] In the formula, Representing velocity space, Indicates linear velocity. Indicates angular velocity, , These represent the minimum linear velocity and the maximum linear velocity, respectively. , These represent the minimum angular velocity and the maximum angular velocity, respectively.

[0101] When the autonomous cable-laying vehicle accelerates or decelerates, the dynamic tension of the cable... It will change accordingly:

[0102] ;

[0103] Among them, the static tension of the cable :

[0104] ;

[0105] Add tension constraint after coupling cable tension. Exceeding this limit will cause the cable to break, resulting in mission failure.

[0106] In summary, the acceleration along the line of motion of the autonomous wire-laying vehicle can be obtained. constraint:

[0107] ;

[0108] In the formula, This indicates the cable tension at the connection point between the cable and the autonomous cable-laying vehicle.

[0109] (2) Modify the dynamic window. The dynamic window of the traditional DWA algorithm Defined as:

[0110] ;

[0111] in, For maximum linear acceleration, For the maximum linear deceleration, Indicates the sampling time interval.

[0112] This embodiment takes into account cable tension constraints and modifies the maximum linear acceleration as follows:

[0113] ;

[0114] Improved dynamic window for:

[0115] ;

[0116] In the formula, This represents the maximum linear acceleration after considering cable tension constraints. This indicates the angular velocity of the autonomous wire-laying vehicle. This represents the maximum angular acceleration after considering cable tension constraints.

[0117] (3) Update speed sampling range.

[0118] Traditional velocity sampling ranges are based on kinematic constraints and dynamic windows. With coupled tension, the tension constraints need to be additionally checked during sampling for each sampling velocity pair. Calculate the corresponding and And only if the condition is satisfied At that time, the calculated and Only then will it be retained and used.

[0119] (4) Modify the trajectory calculation formula.

[0120] Traditional trajectory prediction is based on a uniform velocity model:

[0121] ;

[0122] Considering the tension constraint, the uniform acceleration model is more accurate. In this embodiment, the trajectory calculation formula is modified to a piecewise uniform velocity model under cable tension constraint:

[0123] ;

[0124] ;

[0125] In the formula, express The linear velocity of motion at any given moment. express of Axial component, express of Axial component, express The heading angle at any moment, express angular velocity at any given moment express The linear velocity of motion at any given moment. express of Axial component, express of Axial component, express The heading angle at any given moment.

[0126] (5) Update the comprehensive evaluation function.

[0127] The comprehensive cost function of the traditional DWA algorithm for:

[0128] ;

[0129] In the hybrid path planning algorithm, the global path calculated by the JPS-PLUS algorithm serves as the guiding parameter for local path optimization in DWA, forming the global path evaluation sub-function. After coupling the cable tension model, a tension cost evaluation function is defined. :

[0130] ;

[0131] in, This represents the tension weighting coefficient. This indicates the cable tension safety threshold.

[0132] The new comprehensive evaluation function in this embodiment :

[0133] ;

[0134] In the formula, This represents the heading angle evaluation function. This represents the obstacle distance evaluation function. This represents the current velocity evaluation function. This indicates the global path evaluation sub-function. This represents the tension cost evaluation function. , , , , These represent the weight values ​​of the corresponding evaluation functions. This represents the normalization factor.

[0135] 2.3 Self-rescue guarantee.

[0136] During the movement of the autonomous power line vehicle, new obstacles may appear, blocking the original desired route. If the obstacle is a simple object and does not completely block the passable area, the autonomous power line vehicle can avoid obstacles and find its way using a local path optimization algorithm. However, if the new obstacle is a "U" or "L" shaped obstacle (such as a dead end), or if a new obstacle fills a position that was not originally a "U" or "L" shaped obstacle, creating a "U" or "L" shaped obstacle, the original global guidance path will become invalid, and the autonomous power line vehicle will be unable to continue guiding itself towards the target point. The autonomous power line vehicle will then get stuck in a new local minimum. Therefore, self-rescue safeguards, including an "escape strategy" and a "recovery mechanism," are added to the local path optimization algorithm.

[0137] (1) Escape strategy: Assume the previous guiding path node is The next guide path node is The distance between the two guide path nodes is Algorithm settings: When the autonomous wiring vehicle departs from... Towards If you are moving forward and have not yet arrived In this case, the distance has already exceeded twice the normal distance. The distance of the length indicates that the current working condition is that the autonomous wire-laying vehicle has fallen into a local minimum. The mission safety finite state machine switches from the S3 autonomous wire-laying vehicle obstacle avoidance and path finding state to the S4 self-rescue state and starts the "escape strategy".

[0138] In state S3, the comprehensive evaluation function This includes: heading angle assessment, obstacle distance assessment, current speed assessment, global tracking path assessment, and cable tension assessment. In S4 state, the comprehensive evaluation function... It only includes: obstacle distance assessment, current speed assessment, and cable tension assessment.

[0139] ;

[0140] in, This represents the obstacle distance evaluation function in a self-rescue state. This represents the velocity evaluation subfunction under self-rescue conditions. , , These represent the weight values ​​of the corresponding evaluation functions under the self-rescue state.

[0141] In self-rescue mode, the obstacle distance evaluation sub-function The score is no longer based on how far away from the obstacle, but rather on how close one gets to the minimum safe distance. The higher the obstacle distance assessment value, the faster the speed; the speed assessment subfunction... The higher the value, the more likely the actual escape behavior will be: moving forward close to the edge of the obstacle. This allows the autonomous wiring vehicle to move along the edge of the obstacle, rather than remaining stuck in place.

[0142] Activate the "escape strategy" and use Simultaneously, the original comprehensive evaluation function for each sampling point is recorded. The resulting evaluation value When it appears ,illustrate The new predicted trajectory obtained at time t is already better than The time trajectory has a higher comprehensive evaluation value. The autonomous wire-laying vehicle has successfully escaped the predicament. The mission safety finite state machine has switched from state S4 to state S3, and the self-rescue has ended.

[0143] (2) Recovery mechanism.

[0144] When the mission safety state machine enters state S4, it initiates the "escape strategy" while simultaneously monitoring its effectiveness. If the state machine fails to transition to state S3 within a prolonged period, to ensure the mission's early completion, the "escape strategy" is deemed invalid, and a "recovery mechanism" is immediately activated. After recording important system operation data, the system switches from state S4 to state S1, using the current environmental data as parameters to re-plan the global path. Specifically, when the autonomous wiring vehicle gets stuck in a local minimum, the "escape strategy" is activated, and the current coordinates are recorded. Simultaneously calculate the previous guiding path node. With the next bootstrap node distance When the actual distance traveled exceeds twice the speed again And again (If the coordinate error is set to ±0.1 map units), it is determined that the "escape strategy" has failed, the autonomous wiring vehicle cannot get out of trouble, the "recovery mechanism" is activated, the current working condition information is recorded, and the system enters the S1 state.

[0145] When an obstacle appears that is very small compared to the size of the current task environment map, the task decision layer enters the S4 state and activates the "escape strategy." This avoids increasing the computational load on the algorithm planning layer and the task completion time due to a large map size and multiple calls to the global path planning algorithm. The "recovery mechanism" prioritizes task safety over path quality, ensuring the safe and continuous execution of the task.

[0146] 3. Terminal execution layer simulation experiment.

[0147] This embodiment analyzes and verifies the performance of the hierarchical collaborative control framework's task control logic and hybrid path planning algorithm by simulating the terminal's wiring operation process. It is assumed that the autonomous wiring vehicle in the experiment is equipped with a computing unit and sensing sensors. The map sizes are divided into small (20m×20m), medium (50m×50m), and large (100m×100m), and the obstacle density is divided into low density (10% obstacle density) and high density (20% obstacle density). The autonomous wiring vehicle's speed is measured in m / s. Initially, the autonomous wiring vehicle's maximum speed is 0.8 m / s, and its maximum acceleration is 0.2 m / s². 2 The maximum angular acceleration is (rad / s) 2 The cable laying speed is highly consistent with the speed of the autonomous cable-laying vehicle. The cable acceleration response time is 0.5s, the maximum allowable tension of the cable is 500N, the safe tension is 400N, the laying control gain is 1.2 times, and a 5m long catenary segment is reserved in the initial state. A simulation experiment is designed.

[0148] (1) JPS-PLUS algorithm performance analysis: The test selected 60 standard two-dimensional grid maps with different sizes and obstacle densities as the benchmark. The starting point of all maps was (0.5, 0.5) and the ending point was the top right corner of the map. According to different sizes, they were (19.5, 19.5), (49.5, 49.5), and (99.5, 99.5). The 60 maps include: 10 small-sized (20m×20m) low-density (10% obstacle density) maps, 10 small-sized (20m×20m) high-density (20% obstacle density) maps, 10 medium-sized (50m×50m) low-density (10% obstacle density) maps, 10 medium-sized (50m×50m) high-density (20% obstacle density) maps, 10 large-sized (100m×100m) low-density (10% obstacle density) maps, and 10 large-sized (100m×100m) high-density (20% obstacle density) maps. Comparative tests were conducted on these maps, running both the traditional JPS algorithm and the JPS-PLUS algorithm of this embodiment. The experimental results are as follows: Figures 3-8 As shown in the diagram, black squares represent obstacles, white squares represent passable areas, green squares represent the starting point, purple squares represent the ending point, the yellow broken line represents the overall path, red dots represent explored nodes, and green dots represent nodes yet to be explored. It is evident that the JPS-PLUS algorithm (such as...) Figure 4 , Figure 6 , Figure 8 ) will not occur as Figure 3 , Figure 5 , Figure 7The traditional JPS algorithm directly traverses diagonally adjacent obstacles for pathfinding. The total number of nodes to be explored is equal to the sum of the number of red dots (explored nodes) and the number of green dots (unexplored nodes). The JPS-PLUS algorithm requires significantly fewer nodes to explore than the traditional JPS algorithm, especially with larger map sizes where the optimization is more pronounced. This improvement greatly enhances the algorithm's computational efficiency. Tests on three maps of varying complexity demonstrate that the JPS-PLUS algorithm adapts well to diverse complex environments.

[0149] This embodiment uses 60 maps of varying complexity for experimental comparison. Specific comparison metrics include five indicators: the number of path nodes planned on all maps, the number of explored nodes, the number of nodes to be explored, the path length, and the algorithm's computation time. The detailed data comparison results for each indicator are shown in Table 4. In Table 4, J represents the traditional JPS algorithm, and JP represents the JPS-PLUS algorithm.

[0150] Table 4 Comparison of Algorithm Metrics on Maps of Different Complexities

[0151]

[0152] As shown in Table 4, the JPS-PLUS algorithm in this embodiment shows the most significant improvement in runtime compared to the traditional JPS algorithm, especially on large-size, high-density maps, achieving an improvement of 87.90%. In all test scenarios, the algorithm runtime is improved by more than 30%, indicating that the larger the map and the higher the obstacle density, the more significant the optimization improvement. Regarding node search efficiency, the total number of explored nodes is reduced by 30%-80%, the total number of unexplored nodes is reduced by 17%-66%, and the number of path nodes is reduced by 51%-74%, significantly reducing the cost of maintaining the node list. The number of path nodes ultimately forming the path is also significantly reduced, indicating that the JPS-PLUS algorithm in this embodiment effectively improves path smoothness and reduces path inflection points. The improvement in path length optimization fluctuates within ±2%, with a slight increase in some scenarios, but remains generally controllable, indicating that optimizing the algorithm's runtime efficiency has minimal negative impact on path quality, ensuring the finding of the ideal optimal path. As shown in Table 4, compared to scenarios with larger scale and higher obstacle density, the JPS-PLUS algorithm is more suitable for path planning tasks in large-scale and highly complex environments.

[0153] (2) Performance analysis of hybrid path planning algorithm for coupled cable tension model.

[0154] The test was conducted to determine whether the autonomous wiring vehicle could complete the pathfinding and wiring tasks strictly under the guidance of the hybrid path planning algorithm, provided that the map information remained unchanged during the simulation. Figure 9 and Figure 11The maps used were all small-sized (20m × 20m) maps with low and high obstacle densities (10% obstacle density and 20% obstacle density respectively). The autonomous cable-laying vehicle's operational data is shown in Table 5. It can be seen that, under global path guidance, the autonomous cable-laying vehicle plans the shortest path suitable for its movement, while simultaneously monitoring cable length and tension changes to ensure tension remains within safe thresholds. The autonomous cable-laying vehicle operates at low obstacle density (... Figure 10 ) and high obstacle density ( Figure 12 The average speeds on the map are 0.7655 m / s and 0.7659 m / s, respectively.

[0155] Table 5. Operational data of the autonomous cable-laying vehicle during conventional route finding and cabling.

[0156]

[0157] (3) Experiment on the obstacle avoidance, path finding and wiring capabilities of autonomous wiring vehicle.

[0158] When map information is updated during simulation, the ability of the hybrid path planning algorithm to avoid newly appearing non-moving obstacles is analyzed in maps with the same obstacle density but different sizes. Figure 13 and Figure 15 The obstacle density was low (10%), and the map sizes were small (20m×20m) and medium (50m×50m). The autonomous cable-laying vehicle's operating data is shown in Table 6. It can be seen that under global path guidance and cable tension constraints, the control system can successfully detect new obstacles near the path. Figure 13 and Figure 15 (The red squares in the image) complete map updates and guide the autonomous power line vehicle to avoid new obstacles. The autonomous power line vehicle operates in a small size ( Figure 14 ) and medium size ( Figure 16 The average speeds on the map are 0.7663 m / s and 0.7756 m / s, respectively.

[0159] Table 6. Operational data of the autonomous wiring vehicle during obstacle avoidance, pathfinding, and wiring.

[0160]

[0161] (4) Experiment on the effectiveness of self-rescue guarantee.

[0162] This study analyzes the ability of hybrid path planning algorithms to escape from "L-shaped" and "U-shaped" obstacles (or obstacle zones) and to recover computation when escape fails. Figure 17 and Figure 19 All maps are small-sized (20m×20m) and high-density (obstacle density 20%) maps. The autonomous line-building vehicle operation data is shown in Table 7. Figure 17 and Figure 19The red squares in the diagram represent new obstacles.

[0163] Figure 17 and Figure 18 To test the "escape strategy," an "L-shaped" obstacle was added to the map in real time. The control system successfully detected the new obstacle near the global guidance path. After updating the map information, the obstacle became a "U-shaped" obstacle. At this point, the global path could no longer guide the autonomous cable-laying vehicle forward. Figure 18 Within a range of approximately 200 to 600 steps, the speed of the autonomous wire-laying vehicle was almost 0 m / s for a considerable period, indicating that the autonomous wire-laying vehicle had fallen into a stalemate of a local minimum, and subsequently the "escape algorithm" was invoked. Figure 17 As can be seen, the autonomous cable-laying vehicle did indeed search and advance closely along the obstacles, and finally successfully escaped at a position of approximately (9.5, 10.5), returning to the guidance of the global path and eventually reaching the target point.

[0164] Figure 19 and Figure 20 This experiment, a test of the "recovery mechanism," examines the control system's ability to resume computation when the "escape strategy" proves unsolvable. The results show that because the "recovery mechanism" involves replanning the global path, the initial global guidance path disappears, revealing a new global guidance path as shown below. Figure 19 As shown by the red broken line. (By...) Figure 19 and Figure 20 As can be seen, after successfully detecting new obstacles near the initial global guidance path and updating the map information, the autonomous wiring vehicle's speed remained almost 0 m / s for a considerable period within a range of approximately 300 to 500 steps. This indicates that the autonomous wiring vehicle had entered a stalemate due to a local minimum. At this point, the "recovery mechanism" was invoked, and the hybrid path planning algorithm planned a new global guidance path for the autonomous wiring vehicle. Figure 19 As can be seen from the blue curve (the trajectory of the autonomous wiring vehicle), the autonomous wiring vehicle made a U-turn under the guidance of the new global path, returned to the normal route, and finally reached the target point.

[0165] Table 7. Operational data of the autonomous wire-laying vehicle during the self-rescue test.

[0166]

[0167] In summary, the experimental results show that the hybrid path planning algorithm at the algorithm planning layer not only plans the safest and optimal path for the autonomous wiring vehicle, but also ensures the safety of cable tension during pathfinding and wiring by adjusting the laying speed and the autonomous wiring vehicle's speed. For newly appearing non-moving obstacles, the algorithm can guide the autonomous wiring vehicle to avoid the obstacle with minimal path cost. When a newly appearing obstacle traps the autonomous wiring vehicle, causing the hybrid path planning algorithm to fail, the task decision layer will lead the system into a self-rescue state. This can be achieved by either using an "escape strategy" to ensure the autonomous wiring vehicle can escape when the path calculation gets stuck in a local minimum, or by using a "recovery mechanism" to provide recovery calculation capabilities when the "escape strategy" fails, planning a new global path based on the current map information to guide the autonomous wiring vehicle to continue its task. The hierarchical collaborative control system provides reliable full-process closed-loop control for the mission safety of the autonomous wiring vehicle.

[0168] This invention addresses the lack of mission safety assurance in path planning for autonomous power line erecting vehicles in field emergency scenarios. It establishes a hierarchical collaborative control system of "mission decision-making - algorithm planning - terminal execution," and designs a mission safety finite state machine for the mission decision-making layer. To address the complex operating environment and pathfinding limitations imposed by cable tension safety by autonomous power line erecting vehicles, a hybrid path planning algorithm is designed. The global path plan calculated by the JPS-PLUS algorithm is used as parameters to guide the DWA algorithm coupled with the cable tension model to complete local path optimization. A real-time obstacle recognition module enables avoidance of non-moving obstacles. Experimental results show that the hierarchical collaborative control system and the hybrid path planning algorithm effectively ensure the safety and continuity of autonomous power line erecting vehicle operations, and can provide a technical reference for path planning and mission safety control of intelligent support equipment for field emergency communications.

[0169] Of course, the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A hierarchical collaborative control system for path planning of an autonomous overhead line vehicle, oriented towards mission safety, characterized in that: It includes a hierarchically coordinated task decision layer, algorithm planning layer, and terminal execution layer; The task decision-making layer uses the task safety finite state machine as its core to standardize the division and switching logic of the operation state, realize the identification and self-rescue of abnormal working conditions, and ensure the continuity of task operation. The algorithm planning layer integrates the JPS-PLUS global path planning algorithm, the DWA local path optimization algorithm coupled with the cable tension model, and the YOLO-based real-time obstacle recognition module, and sets the cable tension safety threshold to complete the safe path calculation task. Compared to the traditional JPS algorithm, the JPS-PLUS global path planning algorithm adds detection of diagonal obstacles and pruning of redundant nodes in the path. Compared to the traditional DWA algorithm, the DWA local path optimization algorithm based on the coupled cable tension model: Modify the motion constraints, i.e., add tension constraints after coupling cable tension: ； In the formula, This indicates the acceleration of the moving line of the autonomous wire-laying vehicle; Indicates the maximum allowable tension of the cable; This indicates the cable tension at the connection point between the cable and the autonomous cable-laying vehicle. This indicates the density of the cable's own material; Represents gravitational acceleration; This indicates the length of the suspended section of the cable from the connection point on the autonomous cable-laying vehicle to the ground contact section; Indicates the quality of the independently developed wire-laying vehicle; This indicates the vertical height of the suspended section, which is the height of the connection point between the cable and the autonomous cable-laying vehicle from the ground. Indicates the current total cable length; Indicates the minimum cable length; Modify dynamic windows: ； In the formula, This indicates the improved dynamic window. Indicates linear velocity. Indicates angular velocity, This indicates the linear velocity of the autonomous wire-laying vehicle. This represents the maximum linear acceleration after considering cable tension constraints. This indicates the angular velocity of the autonomous wire-laying vehicle. This represents the maximum angular acceleration after considering cable tension constraints. Indicates the sampling time interval; Update speed sampling range: For each sampling rate pair Calculate the corresponding and And only if the condition is satisfied At that time, the calculated and Only then was it retained for use; ； ； In the formula, Indicates the dynamic tension of the cable. Indicates the static tension of the cable. Indicates the acceleration of cable laying; The terminal execution layer is responsible for obstacle avoidance and pathfinding according to the trajectory issued by the algorithm planning layer, safe wiring, and real-time feedback.

2. The hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety as described in claim 1, characterized in that, The cost function of the JPS-PLUS global path planning algorithm : ； ； ； ； In the formula, Indicates the distance from the starting point to the current node. The actual cost, Indicates the current node Heuristic cost estimation to the destination This represents the straight-line distance from the current node to the destination. This represents the straight-line distance from the starting point to the ending point. This represents the density of obstacles in the rectangular region formed by the current node and the target node. This indicates the number of obstacles within the rectangular area formed by the current node and the target node. This indicates the size of the rectangular region formed by the current node and the target node. The vector consisting of the starting point and the target node. The vector formed by the starting point and the current node cross product, Representing vectors with vector The angle between them.

3. The hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety as described in claim 2, characterized in that, The method for pruning redundant nodes in a path is as follows: determine whether there are obstacle grids in the grid traversed by the line connecting two nodes; if not, delete all redundant nodes between the two nodes. Assuming the node array of the storage path is Define a direct line-of-sight detection function between two nodes. and other node detection functions between two nodes : ； ； In the formula, For the node with the smaller array index, For nodes with larger array indices, , .

4. The hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety as described in claim 3, characterized in that, The steps for direct line-of-sight detection between two nodes and detection of other nodes are as follows: S1, Initialization, Let , ; S2, Connection Node , Generate line segments ; S3, Calling the function If the return value is 1, it means there are other nodes between the two nodes, proceed to step S4; if the return value is 0, it means there are no other nodes between the two nodes, obtain the pruned path and update it. Proceed to step S6; S4. Calling the function If the return value is 1, it means that the two nodes are not directly visible to each other, and the line segment... After passing through the obstacle, proceed to step S6; if the return value is 0, it means that a line segment can be directly connected between the two nodes. If no obstacle is traversed, delete the path node array. , All nodes between, get Let the number of nodes , Proceed to step S5; S5, Order , :like To obtain the path after pruning End of testing; Otherwise, proceed to step S3; S6, Order , Proceed to step S3.

5. The hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety as described in claim 4, characterized in that, Compared to the traditional DWA algorithm, the DWA local path optimization algorithm based on the coupled cable tension model: The trajectory calculation formula has been modified to a piecewise uniform velocity model under cable tension constraints: ； ； In the formula, express The linear velocity of motion at any given moment. express of Axial component, express of Axial component, express The heading angle at any moment, express angular velocity at any given moment express The linear velocity of motion at any given moment. express of Axial component, express of Axial component, express The heading angle at any given moment; Update the comprehensive evaluation function: ； ； In the formula, This represents the comprehensive evaluation function. This represents the heading angle evaluation function. This represents the obstacle distance evaluation function. This represents the current velocity evaluation function. This indicates the function for tracing the global path evaluation. This represents the tension cost evaluation function. , , , , These represent the weight values ​​of the corresponding evaluation functions. Represents the normalization factor. This represents the tension weighting coefficient. This indicates the cable tension safety threshold.

6. The hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety as described in claim 5, characterized in that, The operation process of the mission safety finite state machine is divided into 5 states: S0, S1, S2, S3 and S4. Among them, S0 represents standby, S1 represents global path planning, S2 represents local path optimization, S3 represents autonomous wireline vehicle obstacle avoidance and path finding, and S4 represents self-rescue. When the autonomous line-building vehicle is in S0, if it detects "receiving operation instructions", it will execute "loading map and line-building task information" and transition to S1. When the autonomous wiring vehicle is in S1, if it detects "no reachable path", it will execute "warn that the task has ended abnormally and request clearing", and transition to S0; if it detects "there is a reachable path", it will execute "use the global path as a parameter of the evaluation sub-function to guide local path optimization", and transition to S2. When the autonomous cable-laying vehicle is in S2, if it detects that "the working condition and path calculation are normal", it will execute "set the cable tension safety threshold to guide the autonomous cable-laying vehicle to find the path" and jump to S3; if it detects that "the path calculation is trapped in a local optimum", it will execute "mark the location information trapped in the local optimum on the map" and jump to S4. When the autonomous wiring vehicle is in S3, if it detects "reached the destination", it will execute "mark the mission as completed normally" and jump to S0; if it detects "destination not reached, not trapped in a new obstacle", it will execute "continue to search for a path" and jump to S3; if it detects "destination not reached, trapped in a new obstacle", it will execute "mark the location of the new obstacle on the map and activate the 'escape strategy'" and jump to S4. When the autonomous wiring vehicle is in S4, if it detects "successful escape", it will execute "continue pathfinding" and jump to S3; if it detects "failed escape", it will execute "mark new obstacle locations on the map and activate the 'recovery mechanism'" and jump to S1.

7. The hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety as described in claim 6, characterized in that, When the autonomous wire-laying vehicle is in S4 self-rescue mode, it activates the "escape strategy" and comprehensively evaluates the function. : ； in, This represents the obstacle distance evaluation function in a self-rescue state. This represents the velocity evaluation subfunction under self-rescue conditions. , , These represent the weight values ​​of the corresponding evaluation functions under self-rescue conditions; In a self-rescue state, the closer you get to the minimum safe distance... Obstacle distance evaluation subfunction The higher the value, the faster the speed; the speed evaluation subfunction. The higher the value, the more the autonomous wire-laying vehicle moves along the edge of the obstacle; Activate the "escape strategy" using Simultaneously, the original comprehensive evaluation function for each sampling point is recorded. The resulting evaluation value When it appears ,illustrate The new predicted trajectory obtained at time t is already better than The time trajectory has a higher comprehensive evaluation value. The autonomous wiring vehicle has successfully escaped the predicament. The mission safety finite state machine has switched from state S4 to state S3, and the self-rescue has ended.

8. The hierarchical collaborative control system for autonomous overhead line vehicle path planning oriented towards mission safety as described in claim 7, characterized in that, When the autonomous wire-laying vehicle gets stuck in a local minimum, it activates an "escape strategy" and records the current coordinates. Simultaneously calculate the previous guiding path node. With the next bootstrap node distance ; When the actual distance traveled exceeds twice the speed again And again If the "escape strategy" fails, the "recovery mechanism" is activated, the current operating condition information is recorded, and the system enters the S1 state.

Images