Logistics vehicle path planning method and motion control system

Abstract

The invention discloses a logistics vehicle path planning method and a motion control system, and belongs to the artificial intelligence technology, and the method comprises the steps: initializing map information, and obtaining a starting node coordinate, an obstacle node coordinate and a target node coordinate of a logistics vehicle at present; circularly executing the following steps until a target node is found, and generating an optimal path from the starting node to the target node; executing a self-adaptive search strategy guided based on a target direction from the current node to search candidate nodes; when a jump point exists in the candidate nodes, executing a key jump point security policy improved based on a JPS algorithm to screen security nodes; and through an A * algorithm cost function improved based on environment information and a reward function, an optimal node is screened out from candidate nodes or security nodes to serve as a current node for next search. The method can solve the problems that an A * algorithm of an existing path planning method is not high in search efficiency, and a JPS algorithm is not high enough in safety.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of artificial intelligence, and particularly relates to a logistics vehicle path planning method and a motion control system. BACKGROUND

[0002] With the development of intelligent manufacturing and smart logistics, mobile robots are increasingly widely used in fields such as warehousing, construction, inspection and measurement. The performance of the path planning and control system of the logistics vehicle, as a typical mobile operation platform, directly determines the operation efficiency, safety and task completion accuracy. In a complex environment, achieving high-precision and autonomous path execution control is a key challenge faced by current logistics vehicle control systems.

[0003] Currently, in the field of path planning, global path planning methods are widely used in mobile robot navigation, and the core thereof relies on static map information. Common algorithms include Dijkstra, A*, RRT and jump point search (JPS). Among them, the A* algorithm has good pathfinding ability and interpretability, but in a complex environment, there are problems such as many redundant nodes, dense turning points, and a path that is not smooth enough, resulting in low execution efficiency and high energy consumption of the planned path. The JPS algorithm, as an accelerated improved version of the A* algorithm, reduces the number of path nodes through a jump point strategy, improving search efficiency, but its generated path is too dependent on obstacle boundaries, often hugging obstacles, lacks path safety redundancy, and has problems such as unstable path selection and poor scalability in irregular maps. In addition, the existing path planning system lacks flexibility in adjusting the node expansion strategy, path safety evaluation and dynamic evaluation mechanism, making it difficult to flexibly adjust the heuristic function and path strategy according to environmental changes, limiting its practicality in complex industrial scenarios. SUMMARY

[0004] The present application aims to overcome the deficiencies in the prior art and provide a logistics vehicle path planning method and motion control system, solving the problem of low search efficiency of the A* algorithm and insufficient safety of the JPS algorithm in the existing path planning method.

[0005] To solve the above technical problems, the present application is implemented by using the following technical solutions:

[0006] In a first aspect, the present application provides a logistics vehicle path planning method, comprising:

[0007] initializing map information and obtaining the coordinates of a starting node, obstacle nodes and a target node where the logistics vehicle is currently located;

[0008] cyclically executing the following steps until the target node is found to generate an optimal path from the starting node to the target node:

[0009] searching candidate nodes from the current node by performing a target direction guided adaptive search strategy;

[0010] when there is a key point in the candidate nodes, performing a key point safety strategy based on an improved JPS algorithm to screen safe nodes;

[0011] screening the optimal node from the candidate nodes or the safe nodes as the current node for the next search by a cost function of an improved A* algorithm based on environmental information and a reward function.

[0012] The aforementioned logistics vehicle path planning method, wherein the initialization map information and obtaining the starting node coordinates, obstacle node coordinates and target node coordinates of the current logistics vehicle comprises: initializing the target area as a grid map, creating an open list for storing nodes to be explored and a closed list for storing nodes with calculated costs, obtaining the starting node coordinates, obstacle node coordinates and target node coordinates of the current logistics vehicle, and putting the starting node into the open list.

[0013] The aforementioned logistics vehicle path planning method, wherein the searching candidate nodes from the current node by performing a target direction guided adaptive search strategy comprises:

[0014] calculating the direction vector from the current node to the target node;

[0015] calculating the direction vector from the current node to eight neighborhood directions;

[0016] calculating the included angle by the direction vector to the target node and the direction vector to each neighborhood direction;

[0017] regarding the node with a neighborhood direction with an included angle less than 90 degrees and not being an obstacle as a candidate node;

[0018] when the number of neighborhood directions with an included angle less than 90 degrees is greater than or equal to two, adding the candidate nodes to a key point list;

[0019] when the number of neighborhood directions with an included angle less than 90 degrees is less than two, returning to the eight-neighborhood search strategy of the A* algorithm to search candidate nodes, and adding the candidate nodes to the key point list.

[0020] The aforementioned logistics vehicle path planning method, wherein when there is a key point in the candidate nodes, performing a key point safety strategy based on an improved JPS algorithm to screen safe nodes comprises:

[0021] The key point is a node meeting any of the following conditions: a node with a mandatory neighbor, a starting point and a target point, and a node with a key point in the horizontal or vertical direction when moving diagonally;

[0022] when detecting that there is a key point in the candidate nodes, performing a key point safety strategy on the key point:

[0023] The optimal jump point is screened by calculating the cost of the jump point through the improved A* algorithm cost function;

[0024] Obstacle detection is performed around the optimal jump point to determine the key jump point. If there is an obstacle around the optimal jump point, the optimal jump point is the key jump point. If there is no obstacle around the optimal jump point, the optimal jump point is added to the closed list as the current node for the next search;

[0025] The key jump point is adaptively searched to screen the candidate node as the expansion node;

[0026] A four-way search strategy is performed on the expansion node to screen the expansion node adjacent to the obstacle and the expansion node connected to the obstacle diagonally as the safety node, and the safety node is put into the jump point list.

[0027] The aforementioned logistics vehicle path planning method, the improved A* algorithm cost function includes:

[0028] The environment impact function is designed based on the number of obstacles to improve the heuristic function of the A* algorithm cost function;

[0029] A distance judgment function is designed to dynamically adjust the distance to the target point, and the weight of the heuristic function is dynamically adjusted;

[0030] The reward function is added considering the influence of the key jump point to guide the selection of the safety node;

[0031] The improved A* algorithm cost function dynamically adjusts the balance between path cost and path safety in the node search process.

[0032] The aforementioned logistics vehicle path planning method, the improved A* algorithm cost function F(n) calculation formula is:

[0033] ,

[0034] In the formula, n represents the node to be calculated; G(n) represents the path cost function of the A* algorithm cost function; P(n) represents the distance judgment function; E(n) represents the environment impact function; H(n) represents the heuristic function; D(n) represents the reward function;

[0035] The calculation formula of the environment impact function E(n) is:

[0036] ,

[0037] In the formula, is the number of obstacles in the grid map; is the total area of the map; is the node coordinate to be calculated; is the target node coordinate;

[0038] The calculation formula of the distance judgment function P(n) is:

[0039] ,

[0040] In the formula, represents the starting node coordinates; represents the Chebyshev distance from the node to be calculated to the target node; represents the Chebyshev distance from the starting node to the target node;

[0041] The calculation formula of the heuristic function H(n) is:

[0042] ,

[0043] The calculation formula of the reward function D(n) is:

[0044] ,

[0045] In the formula, represents the maximum step length of node search, represents the minimum step length of node search; diagonal nodes refer to expansion nodes that expand from the key jump point to the diagonal; horizontal special nodes refer to expansion nodes that expand horizontally from the key jump point; ordinary nodes refer to other nodes other than diagonal special nodes and horizontal special nodes.

[0046] The aforementioned logistics vehicle path planning method, the optimal node is selected from the candidate nodes or the safe nodes as the current node for the next search through the A* algorithm cost function improved based on the environmental information and the reward function, comprising:

[0047] Add the jump point list to the open list;

[0048] Calculate the cost of the nodes in the open list through the improved A* algorithm cost function, select the node with the optimal cost as the optimal node, and add the optimal node to the closed list as the current node for the next search.

[0049] The aforementioned logistics vehicle path planning method, after obtaining the current node for the next search, comprising: optimizing the path of the parent node of the current node, the current node and the current node for the next search through the function:

[0050] When it is judged through the function that there is no obstacle between the current node for the next search and the parent node of the current node, the parent node of the current node for the next search is updated from the current node to the parent node of the current node.

[0051] In a second aspect, the present application provides a logistics vehicle motion control system, comprising a path planning module for executing the logistics vehicle path planning method of claim 1 to generate an optimal path from a starting point to a target point.

[0052] The logistics vehicle motion control system further comprises an environment perception module, a control signal transmission module and a system module.

[0053] The environment perception module is configured to acquire surrounding environment information through sensors and construct a grid map of the target area.

[0054] The control signal transmission module is configured to transmit path information of the path planning module, map data of the environment perception module and control instructions of the system module to the logistics vehicle through ROS communication and Socket communication mechanisms.

[0055] The system module is configured to output motion control instructions using a Move_Base framework based on ROS.

[0056] The logistics vehicle receives and executes the motion control instructions of the system module and feeds back motion states to the system module in real time.

[0057] Compared with the prior art, the present application has the following advantages:

[0058] The logistics vehicle path planning method of the present application introduces an adaptive search strategy based on target direction guidance from the current node based on the traditional A* algorithm, further executes a key jump point safety strategy when a jump point is encountered, and uses an improved A* algorithm cost function to screen the current node for the next search; solves the problems of low search efficiency of the existing path planning method A* algorithm and insufficient safety of the JPS algorithm, and can achieve safe and reliable path planning in a complex environment.

[0059] The logistics vehicle path planning method of the present application also adds path optimization to improve path smoothness and obstacle avoidance rationality, and is superior to existing algorithms in terms of the number of corners, path cost and the number of dangerous nodes, so that the logistics vehicle can achieve safe and reliable path planning in a complex environment.

[0060] The logistics vehicle path planning method of the present application is based on the principle of graph search algorithm, and proposes a multi-strategy adaptive algorithm (AE_Astar_MG, Adaptive Environment Astar Modified Guidance), which is based on the Astar algorithm idea and integrates the JPS algorithm jump point strategy and proposes the following four optimization improvements:

[0061] (1) An adaptive search strategy based on vector information guidance is proposed to strengthen the directionality of node search, optimize the search calculation of redundant nodes and improve the path search efficiency.

[0062] (2) Propose a key jump point safety strategy to improve the safety of path planning and optimize some dangerous nodes generated by the jump point strategy;

[0063] (3) Introduce the environment information and the reward function improved A* algorithm cost function to guide the node to dynamically update the heuristic cost according to the environment information in the path search process, balance the two indexes of optimal path cost and path safety;

[0064] (4) Fusion path optimization idea, optimize the number of turning points, reduce redundant nodes, and improve the path smoothness.

[0065] The logistics vehicle path planning method of the present application executes adaptive search strategy in the safe area without obstacles to quickly search along the target point direction, improves the path search efficiency; In the possible dangerous area, the key jump point safety strategy is executed to screen out the optimal node in the safe node, and the safety of the path search is improved; Adaptive search strategy and key jump point safety strategy both use the A* algorithm cost function based on environment information and reward function to select the current node for next search; Combined with path optimization, reduce redundant nodes and unnecessary path turning, improve the path smoothness; The logistics vehicle path planning method keeps the path optimality while improving the search efficiency and path safety, and is suitable for logistics vehicle path planning in complex environment.

[0066] The logistics vehicle motion control system of the present application, through the combination design of the environment perception module, the path planning module, the control signal transmission module and the system module, makes the logistics vehicle motion control system can execute safe and reliable logistics vehicle path planning on the logistics vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 It is a logistics vehicle path planning method flowchart of embodiment 1 of the present application;

[0068] Figure 2 It is a forced neighbor point schematic diagram of embodiment 1 of the present application;

[0069] Figure 3 It is an adaptive search strategy schematic diagram of a logistics vehicle path planning method of embodiment 1 of the present application;

[0070] Figure 4 It is a key jump point safety strategy flowchart of a logistics vehicle path planning method of embodiment 1 of the present application;

[0071] Figure 5 It is a path optimization schematic diagram of a logistics vehicle path planning method of embodiment 1 of the present application;

[0072] Figure 6A path optimization schematic diagram of a logistics vehicle path planning method of embodiment 1 of the present application;

[0073] Figure 7 A logistics vehicle motion control system framework schematic diagram of embodiment 2 of the present application;

[0074] Figure 8 LakiBeam1 three-dimensional model of embodiment 2 of the present application;

[0075] Figure 9 LakiBeam1 radar point cloud data visualization result schematic diagram of embodiment 2 of the present application;

[0076] Figure 10 IMU structure schematic diagram of embodiment 2 of the present application;

[0077] Figure 11 PID control block diagram of embodiment 2 of the present application;

[0078] Figure 12 ROS system architecture schematic diagram of a logistics vehicle motion control system of embodiment 2 of the present application;

[0079] Figure 13 ROS communication mechanism schematic diagram of a logistics vehicle motion control system of embodiment 2 of the present application;

[0080] Figure 14 move_base framework schematic diagram of a logistics vehicle motion control system of embodiment 2 of the present application. DETAILED DESCRIPTION

[0081] The technical solutions of the present application will be described in detail below through the accompanying drawings and specific embodiments. It should be understood that the specific features in the embodiments and the embodiments of the present application are detailed descriptions of the technical solutions of the present application, and are not limitations of the technical solutions of the present application. In the case of no conflict, the technical features in the embodiments and the embodiments of the present application can be combined with each other.

[0082] The term "and / or" in this paper is only a description of the association relationship between the associated objects, which means that there can be three kinds of relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " in this paper generally represents that the associated objects before and after are a "or" relationship.

[0083] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0084] Embodiment one:

[0085] This embodiment introduces a logistics vehicle path planning method, as shown in Figure 1 , which includes:

[0086] S1: initialize map information, and obtain the starting node coordinates, obstacle node coordinates and target node coordinates of the current logistics vehicle;

[0087] S2: perform adaptive search strategy based on target direction guidance from the current node to search candidate nodes;

[0088] When there are jump points in the candidate nodes, perform a key jump point safety strategy based on the improved JPS algorithm to screen safe nodes;

[0089] Through the cost function of the A* algorithm improved based on environmental information and reward function, the optimal node is screened from the candidate nodes or safe nodes as the current node for the next search;

[0090] S3: loop step S2 until the target node is found, and generate the optimal path from the starting node to the target node.

[0091] The logistics vehicle path planning method of this embodiment adopts AE_Astar_MG algorithm for logistics vehicle path planning, which is based on the Astar algorithm idea and combines the JPS algorithm jump point strategy. The specific construction and implementation process of AE_Astar_MG algorithm is introduced as follows:

[0092] Graph search algorithm is a kind of algorithm that screens nodes, forms paths and finds the optimal path of action sequence to reach the target node in the graph structure (generally grid map). These nodes are used to build open list and closed list, and the optimal path is selected through cost calculation. In the list construction process, environmental space obstacle information will be considered to ensure the safety and reliability of the path.

[0093] Astar (A*) algorithm is a graph search algorithm to find the shortest path in single-source path planning problem. It generally expands eight adjacent nodes each time, and then calculates the path cost of each node according to the evaluation function. It searches the target node with the purpose of the shortest path. It includes the actual cost g from the starting point to the current state, and the estimated cost h from the current state to the end point.

[0094] Astar algorithm considers path cost function and heuristic function Two factors are used for path planning, and the evaluation function is defined as:

[0095] (1)

[0096] Heuristic function The calculation method of the heuristic function affects the search efficiency and optimality of the solution. The commonly used heuristic functions include Manhattan distance, Euclidean distance and Chebyshev distance:

[0097] (1) Manhattan distance, see equation (2):

[0098] (2)

[0099] (2) Euclidean distance, see equation (3):

[0100] (3)

[0101] (3) Chebyshev distance, see equation (4):

[0102] (4)

[0103] Astar algorithm path planning process: first input the map, starting point , target point information, initialize OpenList open list to store the nodes to be accessed and CloseList closed list to store the feasible nodes whose cost has been calculated, then put into OpenList, and as the first search point According to the eight-neighborhood search strategy, put the eight nodes around the current node into OpenList, compare the cost of each node, see equation (1), judge whether the adjacent node of is the target point, if yes, return the path Γ, if not, put the node into CloseList and take the node with the minimum cost in OpenList as the current node, and perform the next iteration until the target point is found .

[0104] Because of the redundancy of Astar algorithm in path search performance, too many useless nodes are added to the OpenList and CloseList lists during the node expansion process. To improve this defect, Harabor et al. proposed JPS algorithm. When searching for a path, it discards a large number of nodes that do not need to be calculated, only defines the jump point and adds it to the open list for iterative path search, reduces the calculation of the evaluation function and the memory consumption of maintaining the list, and the overall search speed of the algorithm is faster. The core of JPS algorithm can be summarized as two rules: neighbor pruning rule and jump point screening rule.

[0105] (1) Define the concept of neighbor point: Let the current node be , The adjacent point of the node is , The parent node of the current node is taken as an example in the eight-neighbor expansion mode.

[0106] (5)

[0107] In formula (5), the cost of the path is , and the inequality indicates that when the parent node is the diagonal or non-diagonal node of the node , the path from to without passing through is greater than or equal to the path cost through . When the node neighborhood exists obstacles and formula (5) is established, that is, the straight line path from the parent node p to the node n is blocked by the obstacle, and the path cost from p to n without passing through x is greater than or equal to the path cost through x, it is indicated that the adjacent point is the forced neighbor of the node , Figure 2 which is the eight forced neighbor conditions in the horizontal, vertical and diagonal directions. The forced neighbor is the neighbor node that must be passed through in a certain direction due to the existence of obstacles.

[0108] (2) Jump point pruning strategy: The essence is that the neighbor pruning rule retains the shortest path node in the neighbor, which reduces the search space while ensuring the optimality of the path. The jump point screening rule includes the node with forced neighbor, the starting point or target point, and the point with jump point in the horizontal or vertical direction when searching diagonally. One of the three conditions is met, which determines that the node is a jump point.

[0109] The framework of JPS algorithm is similar to that of Astar (A*) algorithm, which mainly introduces a new list JpList to store the jump points in the iteration and the core function .

[0110] The embodiment fuses and improves the Astar algorithm idea and the JPS algorithm jump point strategy to construct a multi-strategy adaptive algorithm (AE_Astar_MG, Adaptive Environment Astar Modified Guidance):

[0111] Step S1 includes: initializing the target area as a grid map, creating an open list for storing nodes to be explored and a closed list for storing nodes whose costs have been calculated , obtaining the starting node coordinates, obstacle node coordinates and target node coordinates of the current logistics vehicle, initializing the starting node , the target node and the improved A* algorithm cost function, and putting the first starting node into the list.

[0112] The closed list is used to avoid repeated exploration.

[0113] In step S2, the adaptive search strategy based on target direction guidance is performed to search for candidate nodes from the current node, including:

[0114] calculating the direction vector from the current node to the target node;

[0115] calculating the direction vector from the current node to eight neighborhood directions;

[0116] calculating the included angle between the direction vector to the target node and the direction vector to each neighborhood direction;

[0117] taking the node with a neighborhood direction whose included angle is less than 90 degrees and which is not an obstacle as a candidate node;

[0118] when the number of neighborhood directions whose included angle is less than 90 degrees is greater than or equal to two, adding the candidate nodes to the jump point list;

[0119] when the number of neighborhood directions whose included angle is less than 90 degrees is less than two, returning to the eight-neighborhood search strategy of the A* algorithm to search for candidate nodes, and adding the candidate nodes to the jump point list.

[0120] The logistics vehicle path planning method of the embodiment reduces redundant candidate nodes by executing the adaptive search strategy in an open safe area, directly adds the candidate nodes to the jump point list when the number of neighborhood directions whose included angle is less than 90 degrees is greater than or equal to two, adds the jump point list to the open list, selects the optimal node from the candidate nodes as the current node for the next search through the improved A* algorithm cost function.

[0121] The specific construction process of the adaptive search strategy includes:

[0122] The node expansion strategy of traditional A* algorithm generally adopts four-neighborhood or eight-neighborhood search method, which will lead to expand a large number of useless nodes, affect the search efficiency and the processing effect of large map. Therefore, this paper proposes an adaptive search strategy based on the vector information guidance of target point. Considering the existence of "U" type trap in the actual environment, the jump point condition is introduced to avoid falling into local search and degenerate to eight-neighborhood search.

[0123] As shown in Figure 3 , set the ground Figure X OY coordinate system, the current node coordinate is , the target point coordinate is , and the node to be explored is . The norm of the vector from the current node to the target point and the node to be explored can be expressed as:

[0124] (6)

[0125] (7)

[0126] (8)

[0127] (9)

[0128] In the formula, represents the search direction of the current node, represents the search direction, represents the vector norm to the target point, represents the vector norm to the exploration node.

[0129] According to formulas (8) and (9), the angle can be expressed as

[0130] (10)

[0131] The specific idea of adaptive search strategy is as follows:

[0132] Step one: In the node expansion operation, determine whether more than two jump points are generated from the parent node.

[0133] a. If multiple jump points are executed, it means that the current search is in a clear target-oriented area, and the adaptive search strategy can be continued to calculate value;

[0134] b. If not, by default, the node is in a local optimal environment, and the algorithm degenerates to eight-neighborhood search strategy;

[0135] Step two: find the value less than Nodes in a given direction are to be explored; searches in other directions are skipped.

[0136] In step S2, when a hop point exists among the candidate nodes, a critical hop point security strategy based on the improved JPS algorithm is executed to filter safe nodes, including:

[0137] When a hop point is detected among the candidate nodes, the critical hop point security policy is executed on the hop point:

[0138] The optimal jump point is selected by calculating the cost of the jump point using the improved A* algorithm cost function;

[0139] Obstacle detection is performed around the optimal jump point to determine the critical jump point. If there are obstacles around the optimal jump point, the optimal jump point is a critical jump point. If there are no obstacles around the optimal jump point, the optimal jump point is added to the closed list and used as the current node for the next search.

[0140] Adaptive search is used to filter candidate nodes for expansion nodes based on key jump points;

[0141] A four-way search strategy is used to search the extended nodes (searching the four directions of the extended node: up, down, left, and right). Extended nodes that are adjacent to obstacles and those that are diagonally connected to obstacles are selected as safe nodes and added to the jump point list.

[0142] The logistics vehicle route planning method in this embodiment implements a key jump point safety strategy to screen safe nodes in potentially dangerous areas; the jump point list is added to the open list, and the optimal node is selected from the safe nodes using the improved A* algorithm cost function, which is then used as the current node for the next search.

[0143] The specific process of constructing a critical jump point security strategy includes:

[0144] The JPS algorithm suffers from the problem of closely adhering to obstacle boundaries and has numerous dangerous turning points, making it unsuitable for irregularly shaped robots. Ground leveling robots, due to their tail-mounted leveling mechanism, require careful consideration of path safety given their irregular overall structure. Furthermore, the JPS algorithm's insufficient exploration of nodes significantly limits path selectivity, leading to dangerous paths or the possibility of sacrificing path safety.

[0145] Critical jump point security strategies, such as Figure 4 As shown, an adaptive search strategy is applied to eligible hop points for expansion, and obstacle detection is performed on the expanded nodes. The node costs are calculated and updated to increase path selectivity and security. The specific idea of ​​the security strategy method based on key hop points is defined as follows:

[0146] Step 1: Determine if there is a jump point in this iteration. If there is, proceed to the next step; otherwise, skip this iteration.

[0147] Step two: select the optimal jump point in the single iteration list, detect obstacles around the node, if there are obstacles, consider the node as a key jump point and execute the next step, if not, end the key jump point screening and enter step six;

[0148] "not identified as a key jump point", which means that the current jump point is not a high-risk node, and there is no need to perform subsequent key jump point expansion, security optimization and other steps. Skip steps three to five.

[0149] Step three: adaptive search for key jump points to expand key nodes around the key nodes;

[0150] Step four: four-way search strategy (up, down, left and right) for expanded nodes to screen expanded nodes adjacent to obstacles and expanded nodes connected to obstacles diagonally, and update their heuristic cost;

[0151] Step five: put the expanded nodes except obstacles and nodes already in the open list into the jump point list, if the target point is found, return directly, and take the target point as the current node for the next search;

[0152] Step six: add the jump point list to the open list and select the node with the optimal cost in the open list as the next iteration node.

[0153] In step S2, the optimal node is selected from the candidate nodes or safety nodes as the current node for the next search by the A* algorithm cost function improved based on environmental information and reward function, comprising:

[0154] Add the jump point list to the open list;

[0155] Calculate the cost of the nodes in the open list by the improved A* algorithm cost function, select the node with the optimal cost as the optimal node, add the optimal node to the closed list as the current node for the next search.

[0156] The improved A* algorithm cost function includes:

[0157] Design an environmental impact function based on the number of obstacles to improve the heuristic function of the A* algorithm cost function;

[0158] Design a distance judgment function that dynamically adjusts the distance to the target point to dynamically adjust the weight of the heuristic function;

[0159] Consider the influence of key jump points and add a reward function to guide the selection of safety nodes;

[0160] Make the improved A* algorithm cost function dynamically adjust the balance between path cost and path safety in the node search process.

[0161] The specific construction process of the improved A* algorithm cost function includes:

[0162] The embodiment improves the core evaluation function of the graph search algorithm to balance the path cost and the path safety. The heuristic function is one of the main indicators reflecting the algorithm performance, and is generally The closer the cost is to the actual loss, the more nodes the algorithm expands are on the optimal path. When the number of obstacles in the map increases, the difficulty of the accuracy of the nodes to be explored and the estimated cost also increases accordingly. Therefore, the embodiment improves the heuristic function by introducing an environmental impact function , a reward function considering the influence of key jump points to guide the selection of nodes by the algorithm, dynamically adjusting the heuristic cost in the node search process, and balancing the cost and the selection of safe paths.

[0163] The environmental impact function E(n) is expressed as:

[0164] (11)

[0165] In the formula, is the number of obstacles in the grid map; is the total area of the map; is the coordinates of the node to be calculated; is the coordinates of the target node;

[0166] The distance judgment function P(n) is expressed as:

[0167] (12)

[0168] In the formula, represents the starting node coordinates; represents the Chebyshev distance from the node to be calculated to the target node; represents the Chebyshev distance from the starting node to the target node;

[0169] The calculation formula of the heuristic function H(n) is:

[0170] , (13)

[0171] The calculation formula of the reward function D(n) is:

[0172] (14)

[0173] In the formula, represents the maximum step length of node search, represents the minimum step length of node search; in the embodiment, the grid of the grid map is square, and the maximum step length of node search is The grid side length is the minimum step size for node search; diagonal nodes are extension nodes that extend diagonally from the key jump point; horizontal special nodes are extension nodes that extend horizontally from the key jump point; ordinary nodes are other nodes besides diagonal special nodes and horizontal special nodes.

[0174] For the environmental impact function, when the environmental obstacle rate is high, it is more biased towards Manhattan distance, and vice versa. Overall, when the distance to the target point is closer, the improved heuristic function is more biased towards equation (13), and vice versa. The distance judgment function is designed to consider environmental factors and target point distance factors, and dynamically adjust the weight of the heuristic function during the node expansion process. For the expansion of key jump points that meet the conditions, a reward function is adopted. The heuristic cost of the expanded nodes is updated, and the cost of dangerous nodes is increased. Conversely, for ordinary nodes or nodes with good path smoothness, their costs are appropriately reduced to guide the search to prioritize these paths. The improved A* algorithm cost function enhances environmental adaptability, the accuracy of node heuristic costs, and path safety. The improved A* algorithm cost function F(n) is expressed as:

[0175] (15)

[0176] In the formula, n represents the node whose cost is to be calculated; G(n) represents the path cost function of the A* algorithm; P(n) represents the distance judgment function; E(n) represents the environmental impact function; H(n) represents the heuristic function; and D(n) represents the reward function.

[0177] Step S2 also includes path optimization, removing redundant nodes:

[0178] After obtaining the next search node based on the current node, this includes: accessing the parent node of the current node, the current node, and the next search node through... Function path optimization:

[0179] When passing If the function determines that there are no obstacles between the current node and its parent node in the next search, it will update the parent node of the current node to the parent node of the current node in the next search.

[0180] The specific construction process of path optimization includes:

[0181] The neighborhood expansion method based on grid maps is constrained by the grid shape, which limits the feasible turning angle of the search path to 45° or 90°. This embodiment incorporates consideration of the parent node information of the expanded nodes, such as... Figure 5 As shown, collision detection is performed between the extended node and its parent node, the parent node information of the extended node is updated, and the optimized path is obtained. Comparable paths The cost is reduced by more than 7%, while turning points are eliminated to increase path smoothness.

[0182] by Figure 6 For example, from arrive Traditional path planning algorithms aim to avoid black obstacles. That is, the path of the solid black line contains unnecessary turning points. The original path cost and the optimized path cost are expressed as follows:

[0183] (16)

[0184] (17)

[0185] In the formula, G Describes the path cost function for node n. This represents the distance between two nodes.

[0186] If it exists If the cost of equation (17) is less than that of equation (16), then the node is removed. Based on the information, choose the better path. Replace the original path Simultaneously update nodes The parent node information is The new path obtained The Bresenhan algorithm was used to detect... There are obstacles in the path, so the original path should be retained. The final path is This enables the smoothing of jagged paths, greatly improving the problem of small turning point angles in mobile robots and optimizing the actual path cost.

[0187] The overall process of the AE_Astar_MG algorithm is described in detail, and the pseudocode example of the algorithm is shown in Table 1.

[0188] Table 1. Pseudocode Examples of the AE_Astar_MG Algorithm

[0189]

[0190] As shown in Table 1, the initialization phase corresponds to the Input section of the pseudocode and lines 1-3. This involves initializing map information. and ,set up , Nodes, and the first node Put in The list, and loading environmental obstacle information. .

[0191] Entering the loop: corresponding to lines 5-8, according to The function generates the first node, and then... join in In the middle, initialize the jump point list Security policy parameters for critical jump points .

[0192] Adaptive search strategy: Corresponding to lines 14-18 and 37-40 of the pseudocode, it calculates the angle between vectors based on the target point to guide the search. The search direction.

[0193] Path optimization: Corresponding to pseudocode lines 24-27, judgment... The parent node information of the node, through Function judgment node( The parent node of the node is Nodes) and node( Nodes are The parent node of the node) provides information on obstacles between the two nodes; if no obstacles exist, then... The parent node of the node is updated to Nodes. That is, like... Figure 5 As shown, the yellow solid line indicates path1. Node (np) to Node (nc) to The process of node (n) can be simplified to the red dashed line shown in path2. Node (np) to The number of nodes (n) reduces redundant nodes and removes turning points, increasing path smoothness.

[0194] Critical jump point security policy: corresponding to pseudocode lines 33-47, based on Function return The optimal node in the middle is considered to be the most likely to become... Key jump points of nodes ,right Secondary expansion update, update the cost of expansion nodes and put it into The next round of screening will be conducted for the medium-sized items.

[0195] Final Expansion List, filter the current node for the next iteration Until the target point is found Return to global path .

[0196] Example 2:

[0197] Based on the same inventive concept as Figure 7 Embodiment 2 introduces a logistics vehicle motion control system, which includes a path planning module for performing the logistics vehicle path planning method described in Embodiment 1 to generate a global path from the starting point to the target point.

[0198] The logistics vehicle motion control system of the present embodiment further includes an environment perception module, a path planning module, a control signal transmission module, and a system module.

[0199] The environment perception module is configured to acquire surrounding environment information through sensors and construct a grid map of the target area.

[0200] The path planning module is configured to perform the logistics vehicle path planning method described in Embodiment 1 to generate a global path from the starting point to the target point using the AE_Astar_MG algorithm.

[0201] The control signal transmission module is configured to transmit the path information of the path planning module, the map data of the environment perception module, and the control instructions of the system module to the logistics vehicle through ROS communication and Socket communication mechanisms.

[0202] The system module is configured to output motion control instructions using the Move_Base framework based on ROS.

[0203] The logistics vehicle receives and executes the motion control instructions of the system module and feeds back the motion state to the system module in real time.

[0204] In combination with actual use scenarios and functional requirements, the overall scheme framework as shown in Figure 7 is designed.

[0205] To realize each functional module in the above overall scheme, the hardware system of the logistics vehicle is designed as follows in the present example:

[0206] (1) The overall structure mainly meets the requirements of robot motion driving and sensor arrangement, and the driving model and structure of the logistics vehicle are designed.

[0207] (2) The environment perception scheme is based on the completion of basic functions such as robot positioning, SLAM mapping, and motion planning, and selects three sensors, i.e., laser radar, IMU, and motor (with encoder), to meet the data information needs of the mapping module and the path planning module.

[0208] (3) The control scheme includes an industrial computer and a controller, which are used for software layer ROS and data communication, motor control, etc.

[0209] Wheeled robots are a type of mobile robot that has undergone a large amount of engineering practice testing, and is widely used in warehouse, automation, transportation and other engineering practice fields due to its simple structure, efficient movement and other characteristics. In this example, a wheeled mobile robot is selected as the structural form according to the requirements of indoor scene, functional requirements, cost considerations and other requirements. The wheels can be divided into two-wheel differential model, Ackerman model, Mecanum model and omni-directional wheel model according to the different steering modes. The movement ability, control method and adaptive scene of the four models are shown in Table 2.

[0210] Table 2 Comparison table of chassis models

[0211]

[0212] The logistics vehicle of the present example should have good chassis control, robot in-place rotation ability requirement, stable motion control. Excessive motion ability and complex motion control will increase the uncertainty of robot motion, so it is designed as a differential drive wheeled robot.

[0213] Considering the factors of facing a plane scene, compact structure, stable performance and other factors, the LakiBeam1 of Sharp Intelligent Light is selected in this embodiment as shown in Figure 8 , which is a single-line laser radar composed of a single-line laser module and a rotating mechanism, can scan in a two-dimensional plane, has zero blind area within the detection distance, and has good stability for long-time work in actual work.

[0214] The data transmission method of LakiBeam1 is UDP / IP, the data output format is UDP / USB, and the time source is internal timestamp. The point cloud data information visual result can be as shown in Figure 9 .

[0215] IMU is the most basic inertial navigation element. The selection of IMU can refer to multiple performance parameters. IMU contains a gyroscope and an accelerometer, which measure angular velocity and acceleration changes, respectively. The range of IMU represents the ability to measure the maximum angular velocity and acceleration per second. The larger the range, the better the performance. The zero drift stability reflects the size of this zero drift. The smaller it is, the better the stability of the IMU. Considering the comprehensive accuracy, economic benefit and other aspects, the WHEELTEC N100 inertial navigation is selected in this embodiment, and the structure can be seen in Figure 10 .

[0216] The motor is the power control part of the logistics vehicle, in addition to which odometer data information needs to be obtained.

[0217] The motion control of the logistics vehicle mainly relies on the industrial computer sending motion speed instructions to the controller. In order to ensure the standardization and effectiveness of the serial communication between the industrial computer and the controller, a related communication protocol is designed. The protocol mainly includes vehicle speed, forwarding protocol, initial point setting instructions, etc. The general protocol of the control feedback instruction is shown in Table 3.

[0218] Table 3 General communication protocol

[0219]

[0220] Based on the general protocol in the above table, the data content communication protocol of the vehicle speed is designed. Taking the robot moving forward at a speed of 100 r / min with the left wheel and moving backward at a speed of 100 r / min with the right wheel as an example, the vehicle speed communication protocol of the logistics vehicle is shown in Table 4.

[0221] Table 4 Vehicle speed communication protocol

[0222]

[0223] In the vehicle speed communication protocol, direction 1 and direction 2 respectively represent the left and right wheel speeds, the high bit of the wheel speed is in front and the low bit is in back, and the positive and negative bits of the speed are taken as the inverse code when 0xFF.

[0224] The communication protocol design of the initial point setting is shown in Table 5.

[0225] Table 5 Initial point setting communication protocol

[0226]

[0227] In the initial point setting communication protocol, the main function is to send the initial point of the flat plate to the industrial computer ROS end. Type 01 represents motion control, 02 represents PID parameter control, bits 06-10 are initial point information, among which the heading angle is actually enlarged by 100 times, bit 12 is the scale of the map and the real environment, and bits 13-16 are CRC check bits and end bits.

[0228] After receiving the speed information, the controller inputs the speed as the target speed into the speed regulation control algorithm, and finally outputs the control signal PWM (pulse width modulation) signal to the motor drive to control the motor speed, ensuring that the robot can stably complete the predetermined target.

[0229] The speed regulation control system adopts a PID control algorithm. This control system has good flexibility and can realize ordinary PID regulation, as well as PI control, PD control, etc. The appropriate control type can be selected according to the specific control system. Its high reliability and robustness can well realize the speed control of the ground detection robot. The PID control block diagram is shown in Figure 11 .

[0230] According to the system error e(t), i.e. the difference between the target speed and the current speed, the output formula is:

[0231] (18)

[0232] Since the encoder cannot collect continuous pulse signals, the embodiment needs to use a digital PID method to discretize the time. The digital PID algorithm is divided into position PID and incremental PID. The difference between the two is that the position PID uses the error signal to obtain the control amount of the controlled object, and the incremental PID outputs the increment of the control amount, and through weighted processing, a better control effect is obtained, and the influence range is small. Therefore, the embodiment uses incremental PID for speed closed-loop control, and its expression is:

[0233] (19)

[0234] In formula (19), e(k), e(k-1) and e(k-2) respectively represent the current, last and last deviation, , , represents the incremental output.

[0235] According to the above analysis, the incremental discrete PID control process is: 1. Using the M method to measure the speed, the number of pulses per unit time is used to calculate the motor speed, and the speed deviation is obtained with the target speed of the industrial computer; 2. The speed deviation is input to the P, I and D links for calculation, and the last deviation is saved; 3. The output signal PWM controls the motor drive to complete the closed-loop control.

[0236] ROS (Robot Operating System) is an open-source software framework widely used in robot research and development. It integrates a large number of tools, libraries and protocols, and provides standardized communication interfaces for commonly used functional modules. The architecture of ROS is based on the concepts of nodes, topics and services, and through these basic elements, message passing, data sharing and collaboration between modules can be achieved.

[0237] The system architecture of ROS is shown in Figure 12 The embodiment relies on Ubuntu under the Linux system for the development and construction of the ground detection robot; in the middle layer, various robot-related development middleware are encapsulated, and the TCPROS / UDPROS communication system is encapsulated based on TCP / UDP, and various communication mechanisms are realized through the use of publishing / subscription, client / server and other models. In the application layer, developers call different functional packages in ROS to realize related functions, and ROSMaster maintains the normal operation of each node functional package.

[0238] ​The communication mechanism of ROS is as follows Figure 13 , so that each node presents a loose coupling distribution mode. Its core communication mechanism mainly includes the following three kinds: topic communication (Topic), service communication (Server), and parameter server. Topic communication is an asynchronous communication mechanism, and the publisher (Publisher) and subscriber (Subscriber) establish data communication by publishing and subscribing to specified topics (Topic) through the node manager to realize each functional module; service communication is a synchronous communication mechanism, and the client (Client) sends a request to the server (Server) and is in a blocking state, waiting for the server to respond; the parameter server is used to store and manage global parameters, and the node (Node) interacts with the parameter server by obtaining and setting parameters.

[0239] In addition, ROS also provides rich development components and tools such as Rviz, Gazebo, TF coordinate transformation, QT toolbox, Launch file, etc. to help the simulation, debugging and analysis of the flat detection robot and the environment.

[0240] The software system of the logistics vehicle is mainly built based on the topic communication model of ROS. Topic is the carrier of data information communication between nodes. In this example, the key topics of the logistics vehicle are designed and introduced, as shown in Table 6.

[0241] According to the environmental perception scheme and the control signal transmission scheme, the relevant topics such as \imu, \laisai_infos are designed based on the sensor data, and the functions such as logistics vehicle mapping, path planning, obstacle avoidance, data analysis are realized through node publishing and subscribing; the topic / control_msgs is established for speed control communication between the controller and the industrial computer, realizing the motion control of the logistics vehicle; the key topics / socket_msgs and / j_msgs are designed for client communication, laying the foundation for map transmission, client visualization, motion planning, etc.

[0242] Table 6 Key topic content and function table

[0243]

[0244] The navigation system of the logistics vehicle is the key to the implementation of the path planning function of the logistics vehicle. In this embodiment, the motion planning scheme of the robot is mainly established based on the Move_Base framework of ROS, as shown in Figure 14 , realizing the path planning and motion control of the autonomous navigation module and the control signal transmission module of the robot.

[0245] Main components of Move_Base framework:

[0246] (1) Global_Planner: used to find the optimal path of the robot from the starting point to the target point on the static map.

[0247] (2) Local_Planner: used to avoid obstacles and control the trajectory planning of the robot in real time during the robot's travel. The DWA algorithm is used for local path planning in this embodiment.

[0248] (3) CostMap_2D: used to represent the information of obstacles, free space and non-passable areas around the robot.

[0249] (4) Recovery_Behaviors: including a series of self-recovery strategies of the robot when planning fails, such as re-planning the path, rotating the robot, etc.

[0250] (5) Map_Sever: used to load and publish the map.

[0251] (6) sensor information, tf transformation, etc.

[0252] The Move_Base process framework of the logistics vehicle based on ROS includes: first, motion planning is established on the basis of known map information, and the map construction method in this embodiment includes two kinds: robot SLAM mapping and flat user terminal based on Socket communication transmission map. Then the logistics vehicle publishes the measurement task and target point based on the user point, uses the Move_Base framework to subscribe / goal, / map and sensor necessary topics to realize path planning and achieve logistics vehicle motion control. Finally, the logistics vehicle realizes path planning and movement of the path planning module based on this basic framework.

[0253] The overall control process of the logistics vehicle motion control system includes: first, the environment perception module obtains the surrounding environment information through laser radar, IMU and wheel encoder and other multi-source sensors, and constructs a two-dimensional or three-dimensional map of the target area through SLAM mapping algorithm, providing the environment basis for subsequent path planning. After the map construction is completed, the system enters the path planning module, and the improved AE_Astar_MG algorithm is used for optimal path planning to ensure that the generated path has good smoothness, safety and execution efficiency. Subsequently, the control signal transmission module is responsible for stably transmitting path information, map data and control instructions to the logistics vehicle through ROS communication and Socket communication mechanism. At the logistics vehicle end, the system module realizes motion control based on the Move_Base framework of ROS. Finally, the logistics vehicle adjusts the speed in real time according to the received path planning and control instructions, and accurately completes the task of moving to the target point by the PID controller driving the motor.

[0254] Those skilled in the art will appreciate that embodiments of the application can be readily used as software, hardware, or a combination of software and hardware. In one embodiment, the application can be implemented in software and can be stored on a computer readable medium, which can include random access memory (RAM), read only memory (ROM), magnetic disk or optical disk, or the like. The application can also be implemented as a combination of both software and hardware. In addition, the application can be implemented as a computer program product that can include a computer readable medium having stored computer program code thereon.

[0255] The present application is described in reference to the drawings, which are as follows. Figure One one or more processes and / or blocks Figure One means for carrying out the function specified by the block or blocks.

[0256] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the Figure One one or more processes and / or blocks Figure One one or more blocks that can perform the function specified by the block or blocks.

[0257] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions that are executed on the computer or other programmable apparatus provide steps for implementing the Figure One one or more processes and / or blocks Figure One one or more blocks that can perform the function specified by the block or blocks.

[0258] The above description is only preferred embodiments of the application. It is obvious that those skilled in the art can make some improvements and modifications without departing from the technical principles of the application. These improvements and modifications should also be considered as falling within the scope of the application.

Claims

1. A logistics vehicle route planning method, characterized in that, The method comprises the following steps of: initializing map information, and obtaining a starting node coordinate, an obstacle node coordinate and a target node coordinate of a current location of a logistics vehicle; recursively performing the following steps until the target node is found, and generating an optimal path from the starting node to the target node: performing an adaptive search strategy based on a target direction guide to search for a candidate node from a current node; when there is a jump point in the candidate node, performing a key jump point safety strategy based on an improved JPS algorithm to screen a safe node; screening an optimal node from the candidate node or the safe node as a current node for next search through a cost function of an improved A* algorithm based on environmental information and a reward function.

2. The method of claim 1, wherein, The step of initializing map information, and obtaining a starting node coordinate, an obstacle node coordinate and a target node coordinate of a current location of a logistics vehicle comprises the following steps of:

3. The method of claim 1, wherein, initializing a target area as a grid map, creating an open list for storing nodes to be explored and a closed list for storing nodes with calculated costs, obtaining the starting node coordinate, the obstacle node coordinate and the target node coordinate of the current location of the logistics vehicle, and putting the starting node into the open list. The step of performing an adaptive search strategy based on a target direction guide to search for a candidate node from a current node comprises the following steps of: calculating a direction vector from the current node to the target node; calculating direction vectors from the current node to eight neighborhood directions; calculating included angles through the direction vector to the target node and the direction vectors to the neighborhood directions; regarding a node in a neighborhood direction with an included angle less than 90 degrees and not being an obstacle as a candidate node; when there are two or more neighborhood directions with included angles less than 90 degrees, adding the candidate nodes into a jump point list; 4. The method of claim 3, wherein, when there are less than two neighborhood directions with included angles less than 90 degrees, returning to an eight-neighborhood search strategy of the A* algorithm to search for a candidate node, and adding the candidate node into the jump point list. The step of performing a key jump point safety strategy based on an improved JPS algorithm to screen a safe node when there is a jump point in the candidate node comprises the following steps of: The jump point is a node meeting any one of the following conditions: a node with a forced neighbor, a starting point and a target point, and a node with a jump point in a horizontal or vertical direction when moving diagonally; when it is detected that there is a jump point in the candidate node, performing a key jump point safety strategy on the jump point: calculating a cost of the jump point through an improved A* algorithm cost function to screen an optimal jump point; detecting obstacles around the optimal jump point to determine a key jump point, and if there are obstacles around the optimal jump point, the optimal jump point is the key jump point, and if there are no obstacles around the optimal jump point, the optimal jump point is added into the closed list as a current node for next search; performing adaptive search on the key jump point to screen a candidate node as an expansion node; 5. The method of claim 4, wherein, performing a four-square search strategy on the expansion node to screen an expansion node adjacent to an obstacle and an expansion node connected to the obstacle at a diagonal as a safe node, and putting the safe node into the jump point list. The improved A* algorithm cost function comprises the following steps of: designing an environmental impact function based on the number of obstacles to improve a heuristic function of the A* algorithm cost function; designing a distance judgment function dynamically adjusted according to a distance to the target point to dynamically adjust a weight of the heuristic function; considering an influence of the key jump point to add a reward function to guide selection of the safe node; The improved A* algorithm cost function dynamically adjusts the balance between path cost and path safety in the node search process.

6. The method of claim 5, wherein, The improved A* algorithm cost function F(n) is calculated by the following formula: ， In the formula, n represents a node whose cost is to be calculated; G(n) represents a path cost function of the A* algorithm cost function; P(n) represents a distance judgment function; E(n) represents an environmental influence function; H(n) represents a heuristic function; and D(n) represents a reward function. The calculation formula of the environmental influence function E(n) is as follows: ， In the formula, is the number of obstacles in the grid map; is the total area of the map; is the node coordinate to be calculated; is the target node coordinate; The calculation formula of the distance judgment function P(n) is as follows: ， In the formula, denotes the starting node coordinate; denotes the Chebyshev distance of the node to be calculated to the target node; denotes the Chebyshev distance of the starting node to the target node; The calculation formula of the heuristic function H(n) is as follows: ， The calculation formula of the reward function D(n) is as follows: ， In the formula, represents the maximum step length of node search, represents the minimum step length of node search; diagonal nodes refer to expansion nodes that expand toward the diagonal of the key jump point; horizontal special nodes refer to expansion nodes that expand horizontally toward the key jump point; and ordinary nodes refer to other nodes other than the diagonal special nodes and the horizontal special nodes.

7. The cart path planning method according to claim 5 or 6, characterized by, The optimal node is screened out from the candidate nodes or the safe nodes by the improved A* algorithm cost function based on environmental information and the reward function, and the optimal node is used as a current node for the next search, and the method comprises the following steps: The jump point list is added to the open list. The cost of the nodes in the open list is calculated by the improved A* algorithm cost function, the node with the optimal cost is screened out as the optimal node, and the optimal node is added to the closed list and used as the current node for the next search.

8. The method of claim 3, wherein, After the current node gets the current node of next search, including: the parent node of the current node, the current node and the current node of next search are optimized by the function path When passing If the function determines that there are no obstacles between the current node and its parent node in the next search, it will update the parent node of the current node to the parent node of the current node in the next search.

9. A vehicle motion control system characterized by comprising: The path planning module is configured to execute the logistics vehicle path planning method of claim 1 to generate an optimal path from a starting point to a target point.

10. The material handling vehicle motion control system of claim 9, wherein, Further comprising: An environment perception module, a control signal transmission module and a system module; The environment perception module is configured to acquire surrounding environmental information by using a sensor and construct a grid map of a target area. The control signal transmission module is configured to transmit path information of the path planning module, map data of the environment perception module and control instructions of the system module to the logistics vehicle by using a ROS communication and a Socket communication mechanism. The system module is configured to output a motion control instruction by using a Move_Base framework based on ROS. The logistics vehicle receives and executes the motion control instruction of the system module and feeds back a motion state to the system module in real time.

Images