Unmanned logistics vehicle path planning method based on fusion of improved A* and improved DWA
By combining the improved A* algorithm and dynamic window method, screening key nodes and optimizing the DWA algorithm, the problem of insufficient efficiency and security of traditional algorithms in dynamic environments is solved, and efficient and smooth path planning of unmanned logistics vehicles is achieved, which has improved the The autonomous driving capability of unmanned logistics vehicles.
Patent Information
- Application Number
- CN202411166466.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-29
AI Technical Summary
The traditional A* algorithm has low search efficiency in dynamic or complex environments, and the path is not smooth enough. The DWA algorithm has insufficient problems when dealing with moving obstacles, so it cannot effectively avoid collisions.
Combined with the improved A* algorithm and dynamic window method, through parameter settings and grid map initialization, global path planning is carried out, key nodes are selected and local path planning is used to generate the actual driving path of the unmanned logistics vehicle, and optimized The efficiency and security of the path planning algorithm are achieved.
It significantly improves the efficiency and safety of path planning, reduces the computing burden, optimizes the smoothness and obstacle avoidance capabilities of the path, and enhances the autonomous driving ability of the unmanned logistics vehicle in complex environments.
Smart Images

Figure CN119043359A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of unmanned logistics vehicle path planning, and in particular to an unmanned logistics vehicle path planning method based on the fusion of improved A* and improved DWA. Background Art
[0002] In recent years, the rapid development of artificial intelligence technology has promoted the widespread application of autonomous driving and intelligent path planning. At the same time, e-commerce and the global supply chain have developed rapidly. Unmanned logistics vehicles are widely used in short-distance logistics transportation due to their fast and convenient characteristics. Logistics platforms continue to increase investment in unmanned logistics vehicles, reducing transportation costs by replacing manual labor with unmanned methods; at the same time, unmanned logistics vehicles are also equipped with autonomous planning capabilities, which helps to adapt to complex environments safely and efficiently.
[0003] Path planning is the core link in the autonomous mission of unmanned vehicles. An excellent path planning algorithm must meet the efficient and safe working requirements of unmanned vehicles. Efficiency means that the path planning algorithm can find the best path from the starting point to the end point within a limited time. Safety means that the algorithm can ensure that the unmanned vehicle avoids collisions with other vehicles, pedestrians or obstacles. Safety is the primary condition for autonomous driving of unmanned vehicles, so the path planning algorithm must be able to respond to changes in the surrounding environment in a timely manner and plan a reasonable obstacle avoidance path. Path planning technology is divided into two categories: global static planning and local dynamic planning, which respectively deal with path finding in known environments and real-time obstacle avoidance navigation in unknown or changing environments. Classic algorithms such as A*, artificial potential field method, RRT, and genetic algorithm dominate global planning, while dynamic window method, time elastic band algorithm, etc. show high efficiency in local obstacle avoidance.
[0004] The traditional A algorithm is suitable for global path planning in static environments, but its search efficiency is low in dynamic or complex scenarios, and the path is not smooth enough. Yue Gaofeng et al. tried to improve this situation by combining bidirectional search with cubic Bezier curve smoothing, but often failed to effectively reduce the redundant nodes in the path. Lao Cailian et al. reduced path redundancy by selecting key points. Although this can improve path smoothness and safety, it does not guarantee the shortest path and more visited nodes. Yang Jiahua et al. improved the A algorithm by reducing the number of path turns. Although it improved the trajectory smoothness, it sacrificed the search speed. In terms of local path planning, the DWA algorithm achieves obstacle avoidance by solving the optimal solution in the velocity space. Li Di et al. proposed a method to adjust the evaluation function based on the change of obstacle density, but it is insufficient when dealing with moving obstacles because the turning action is only triggered when approaching the obstacle. Summary of the invention
[0005] In view of the limitations and shortcomings of the prior art, the present invention proposes an obstacle avoidance method for unmanned logistics vehicles that combines an improved A algorithm with a dynamic window method (DWA). The method first sets parameters and initializes the grid map, and clarifies the starting and ending points of the unmanned logistics vehicle. Then, the improved A* algorithm is used for global path planning to obtain a series of continuous path nodes. Key nodes are selected from these path nodes as the targets of local path planning. Finally, the optimized DWA algorithm is used to complete local path planning, thereby generating the actual driving path of the unmanned logistics vehicle.
[0006] Technical means adopted by the present invention
[0007] A path planning method for unmanned logistics vehicles based on the fusion of improved A* and dynamic window method includes the following steps:
[0008] S1. Initialize the parameters and grid map, and set the end point and starting point of the unmanned logistics vehicle path planning;
[0009] S2. Perform global path planning in the initialized map based on the improved A* algorithm to obtain a preliminary global path;
[0010] S3, using the tertiary redundant point deletion strategy to process the nodes in the preliminary global path planning to generate a global optimized path;
[0011] S4. Obtain the key turning points in the optimized global path, and use the key turning points as the target points for local path planning. Then, perform local path planning based on the improved DWA algorithm to obtain the optimal motion trajectory.
[0012] Furthermore, a global path planning is performed in the initialized map based on the improved A* algorithm to form a preliminary global path, including:
[0013] S201, obtaining the starting position, target position and environmental obstacle information of the unmanned logistics vehicle, and then calculating the obstacle density based on the obtained environmental obstacle information and the starting position and target position of the unmanned logistics vehicle;
[0014] S202, starting from the beginning. Use the evaluation function of the improved algorithm to select the best child node in the neighborhood around the parent node. Then use the best node as the parent node to continue searching for subsequent child nodes until the target position is reached. The weight parameter of the evaluation function of the improved A* algorithm is adjusted by a method of dynamically adjusting the weight of the heuristic function estimation value;
[0015] S203, backtracking the optimal node extracted each time to form a preliminary global path.
[0016] Furthermore, the nodes in the initially formed global path are processed using the tertiary redundant point deletion strategy to generate the global optimal path, including:
[0017] S301, the system checks all nodes in the path in sequence, and by determining that the current node is in the same straight line with the two nodes before and after it, the node is considered redundant and deleted from the path nodes;
[0018] S302, after the deletion operation, check the nodes in the path one by one, and define the nodes on the path as a set of sequences {Tk|k=1,23…,n}. First, find the connection point P between the T2T3 nodes from the T1 node. Connect T1 to the middle point P of T2T3, and check whether the path crosses the obstacle area. If so, consider connecting T1 directly to P. If not, find whether there is a point in the middle of T3T4 that can meet the requirements, until a point is found to connect T1, and continue to infer from this node to find the final point TK.
[0019] S303. In this case, a new node is obtained, and the previous step is performed again until the final point TK is found, and all nodes after the redundant points are deleted again are obtained.
[0020] Furthermore, the key turning points in the global optimization path are obtained, and the key turning points are used as local target points. Local path planning is performed based on the improved DWA algorithm to obtain the optimal motion trajectory. Including:
[0021] S401. A kinematic model is established based on the motion of the unmanned logistics vehicle. The model uses linear velocity and angular velocity to calculate the position and posture changes of the smart vehicle in continuous time intervals, in which the heading angle is also taken into consideration. The data obtained through sampling accurately describes the small displacement of the smart vehicle along the straight line, and then simulates countless different drivable trajectory routes;
[0022] S402, using the obtained key turning point as the local target point of the DWA optimization algorithm, using the improved evaluation function to evaluate the quality of the trajectory, and selecting the optimal trajectory;
[0023] S403, introducing a dynamic safety distance of an obstacle, and evaluating the collision risk according to the moving speed of the obstacle. The faster the obstacle moves, the longer the safety distance is, and the slower the obstacle moves, the shorter the safety distance is.
[0024] S404, saving the optimal travel trajectory speed space after real-time planning;
[0025] S405. The unmanned logistics vehicle moves according to the processed optimal trajectory, and after reaching the set local target point, determines whether it has reached the final target point. If it has not reached the final target point, it will return to step S401 and continue the next stage of path planning until the entire navigation task is completed.
[0026] Compared with the prior art, the present invention has the following advantages:
[0027] 1. The present invention reduces the computational burden and the path planning time by optimizing the neighborhood search selection strategy; at the same time, the heuristic function is optimized to dynamically evaluate the obstacle density by calculating the rectangular area between each node subnode and the target point, as well as the total area occupied by obstacles in the area. Combining this density value with the traditional heuristic function can provide richer environmental information for the search algorithm. This method enables the algorithm to give priority to areas with fewer obstacles, thereby avoiding falling into complex local optimal solutions and improving the efficiency and quality of path planning. Finally, the three-time redundant point deletion strategy is used to remove redundant points, leaving key turning points, which shortens the path planning length and improves the smoothness of the path.
[0028] 2. The present invention improves the original DWA algorithm. First, the trajectory sampling evaluation function is optimized to help it get close to the global path planned by the A* algorithm. Secondly, a safety buffer zone around the obstacle is set by adding an obstacle dynamic evaluation submodule. This safety distance can be flexibly adjusted according to the instantaneous speed of the obstacle, following the principle that the safety buffer zone near obstacles with lower speeds can be appropriately reduced. In the path planning stage, the real-time position of the obstacle is accurately evaluated, and the speed and distance factors are comprehensively considered to adopt the corresponding obstacle avoidance strategy.
[0029] 3. The present invention provides an efficient obstacle avoidance algorithm that has both static global obstacle avoidance and dynamic local obstacle avoidance capabilities, aiming to overcome the limitations of traditional A* and DWA algorithms. The algorithm guides local path planning by improving the key nodes extracted by the A* algorithm, and realizes dynamic obstacle avoidance in combination with the optimized DWA algorithm. Compared with the use of the A* or DWA algorithm alone, the present invention significantly improves the efficiency of path planning, the sensitivity and safety of obstacle avoidance, thereby improving the work efficiency and safety reliability of unmanned logistics vehicles in short-distance operations, showing significant practical application value. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0031] Figure 1 This is a flow chart of a path planning method for an unmanned logistics vehicle based on the fusion of improved A* and improved DWA of the present invention.
[0032] Figure 2This is a graph for selecting the moving direction of a search node in an embodiment.
[0033] Figure 3 Graph showing the correspondence between the angle α and five selected directions in the embodiment.
[0034] Figure 4 This is the traditional 8-domain A* search node graph in the embodiment.
[0035] Figure 5 Improve the 5-domain A* search node graph in the embodiment.
[0036] Figure 6 It is a simulation diagram of the traditional A* algorithm route in the embodiment.
[0037] Figure 7 This is a simulation diagram of the improved evaluation function A* algorithm route in the embodiment.
[0038] Figure 8 This is a simulation diagram of the A* algorithm route for three redundant point deletions in the embodiment.
[0039] Fig. 9 This is a simulation diagram of the improved A* algorithm path in the embodiment.
[0040] Fig.10 It is a simulation diagram of the fusion algorithm path of the present invention in the embodiment. DETAILED DESCRIPTION
[0041] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solution in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0042] like Figure 1 As shown, the present invention proposes a path planning method for an unmanned logistics vehicle that combines the improved A algorithm with the DWA algorithm. The overall process is summarized as follows: First, the parameters and grid map are initialized, and the starting position and destination of the unmanned logistics vehicle are set. Next, the improved A* algorithm is used to perform global path planning to generate a series of continuous path nodes. The key nodes are selected as the destinations for local path planning. Finally, the optimized DWA algorithm is used to complete the local path planning, thereby determining the actual driving path of the unmanned logistics vehicle. The specific steps include:
[0043] S1. Initialize the parameters and grid map, and set the end point and starting point of the unmanned logistics vehicle path planning.
[0044] S2. A global path planning method based on the improved A* algorithm is implemented as follows:
[0045] S201, using a heuristic method based on relative density of obstacles, dynamically calculating the weight of the obstacle ratio affecting the heuristic function h(n) according to each child node of the current node, the target point and the known obstacle position.
[0046] S202, calculate the total area s of the rectangular range from one of the child nodes of the parent node to the target point, then calculate the total area t of all obstacles in the rectangle, and obtain the density formula of the obstacle area in the rectangle from the child node to the target point to the total area:
[0047] r=t / s
[0048] S203, introducing the obstacle ratio r into the evaluation function f(n), changing the weight of the heuristic function h(n), and realizing the dynamic adjustment of the evaluation function f(n). The heuristic function h(n) selects the Euclidean distance. The improved evaluation function f(n) formula is as follows:
[0049] f(n)=g(n)+e r *h(n)
[0050] S204, optimizing the node path using the search point selection strategy, the implementation steps are as follows:
[0051] S205, the search principle of the traditional A* algorithm is to add a starting point to the linked list. Then, add the point as a parent node to the closed list. Next, search its adjacent child nodes in 8 directions such as Figure 2 , calculates the possible cost of traveling in any direction from 1 to 8 and selects the node to move to next. When facing different environments, its simple search strategy is rigid, resulting in unnecessary waste of search resources.
[0052] S206. In order to improve the search efficiency, the current node is connected to the target node, and the angle between the current node and the positive semi-axis direction of the Y axis is set to α. The relationship between the α angle and the search direction is as follows: Figure 3 As shown, five search directions are retained and three search directions are discarded.
[0053] S207, the specific process of the improved A* algorithm is as follows:
[0054] Add the starting node to the open list, which is used to store the nodes to be evaluated;
[0055] Enter a loop until the target node is found or the open list is empty:
[0056] Select the node with the lowest f(n) value from the open list as the current node;
[0057] Remove the current node from the open list and add it to the closed list, which is used to record the nodes that have been evaluated;
[0058] Perform the following operations on the five neighboring nodes of the current node:
[0059] If the neighboring node is already in the closed list, it is ignored;
[0060] If the neighboring node is not in the open list, calculate its f(n) value, set its parent node as the current node, and add it to the open list;
[0061] If the neighboring node is already in the open list, check whether the path to it through the current node is better (i.e., the g(n) value is lower). If so, update its f(n) value and set its parent node to the current node;
[0062] The algorithm ends when the target node is added to the closed list;
[0063] Starting from the target node, trace back along the parent node chain to the starting node. This path is the optimal path.
[0064] S3, optimize the path nodes using the triple redundant point removal strategy, the implementation steps are as follows
[0065] S301. For each pair of adjacent nodes, determine whether path optimization can be performed by calculating the angle between their connecting line and the connecting line of the next node. If the two lines are collinear, directly connect the farther node to shorten the path; otherwise, calculate the relationship between the straight line formed by the two connecting lines and the obstacle. Only when there are no obstacles within the straight line range, it is allowed to connect the farther node, otherwise the original path structure remains unchanged.
[0066] S302, a step-by-step refinement strategy is adopted to examine the node relationships in the path pair by pair, and dynamically find the optimal intermediate connection point between each pair of nodes, rather than judging the straightening only based on the angle. For each pair of nodes, we calculate all possible connection points along the line segment and evaluate whether they will cross the obstacle area when forming a line with the current point. If there is an intermediate connection point that meets the obstacle avoidance requirements, it will be included in the new optimized path, otherwise the next node in the original path will be retained.
[0067] 303. Perform the previous step again to achieve recursive optimization, and analyze the node connections in detail to eliminate redundancy and ensure that the path is barrier-free. By re-examining the optimized path, explore the connection options of each pair of nodes in more detail, and only add those relay points that do not affect obstacle avoidance to reduce redundant nodes and improve the smoothness and efficiency of the path.
[0068] S4. Perform local path planning based on the improved DWA algorithm. The implementation steps are as follows:
[0069] S401, initialize the speed, acceleration and evaluation function parameters of the unmanned logistics vehicle, and then calculate a speed range according to the current state of the unmanned logistics vehicle, which is the dynamic window. Within this window, the algorithm will simulate the future position at different speeds and generate multiple possible driving trajectories.
[0070] Due to the performance limitations and environmental factors of unmanned logistics vehicles, the actual speed is affected by its maximum and minimum speed thresholds, motor performance and braking distance. These three conditions together define its reasonable range of motion.
[0071] The maximum and minimum speed constraints are:
[0072]
[0073] Where: v min ,v max is the minimum and maximum linear speed, ω min ,ω max are the minimum and maximum angular velocities.
[0074] Due to the motor performance constraints, the smart car has a maximum and minimum speed due to acceleration, which is:
[0075]
[0076] Where: v c ,ω c is the current speed; v a ,ω a is the maximum acceleration; v b ,ω b is the maximum deceleration.
[0077] To ensure the safety of the smart car, under the condition of maximum deceleration, the current speed should be able to decelerate to 0 before hitting the obstacle, then:
[0078]
[0079] Where: dist(v,ω) is the shortest distance to the obstacle on the trajectory corresponding to (v,ω).
[0080] S402 In summary, the feasible prediction speed is the intersection of the three sets, and the dynamic window feasible speed space is as follows:
[0081] V r =V m ∩V d ∩V a
[0082] S403. Traditional DWA algorithms often have difficulty in effectively avoiding collisions, and even occasionally cause system interruptions. Therefore, we propose to expand the original dynamic window algorithm by integrating the obstacle dynamic assessment submodule and implementing the initial normalization configuration of the weight coefficient, as shown below:
[0083]
[0084] number(v,ω)=x t +y t -μln(x i -1×y i -1)
[0085] Where M is the area of the map; (x s ,y s ) is the starting coordinate of the unmanned vehicle; (x t ,y t ) is the coordinate of the target point of the unmanned vehicle; (x i ,y i ) is the coordinate of the unmanned vehicle at time i; number(v,ω) is the obstacle number function.
[0086] S404. The key turning points extracted from the third part will be used as local target points to guide the local movement direction of the robot. The unmanned logistics vehicle uses this to enhance the algorithm's path planning adaptability in a dynamic environment and the obstacle avoidance safety of the intelligent vehicle. The improved evaluation function is:
[0087] G(v,m)=α×heading(v,ω)+β×dist(v,ω)+γ×velocity(v,ω)+δ×number(v,ω)
[0088] Where, heading(v,ω) is the azimuth evaluation function, which indicates the azimuth deviation between the simulated predicted path end position and the global path node; velocity(v,ω) indicates the speed of the unmanned boat in the current estimation; dist(v,ω) is the distance evaluation function, which indicates the shortest distance between the global known obstacle and the simulated trajectory and the shortest distance between the local unknown obstacle and the simulated trajectory. α, β, γ, and δ are the weight coefficients of the response of each sub-function of the evaluation function.
[0089] S405: retain the optimal trajectory speed space.
[0090] S406. The unmanned logistics vehicle moves along the optimal trajectory and determines whether it has reached the end point after reaching the local target point. If it has reached the end point, the operation ends. If it has not reached the end point, steps S401-S405 are repeated.
[0091] To prove the feasibility of the algorithm proposed in this paper, MATLAB simulation software was used for testing. The simulation environment is a grid map simulating a complex community. The white grid in the map is the feasible domain, the black grid is the obstacle, △ is the starting point, and ○ is the target point. A 25*25 grid map is designed for the working environment of the unmanned logistics vehicle to conduct a simulation comparison experiment. Figure 4 represents the traditional 8-domain A* search node, Figure 5 Represents an improved 5-domain A* search node, Figure 6 Represents the traditional A* algorithm route simulation, Figure 7 The route simulation results of the A* algorithm with improved evaluation function are shown in Figure 2. Figure 8 It represents the simulation result of the A* algorithm route with three redundant points removed. Fig. 9 The figure shows the simulation results of all improved A* algorithm routes. Both algorithms successfully plan a complete and feasible global path, but the performance of the planned paths is different. The specific experimental parameters are shown in the following table.
[0092] Table 1 Comparison experimental parameters of two A* algorithms
[0093]
[0094] By analyzing the data in Table 1, it can be seen that the improved A algorithm effectively narrows the search range by optimizing the search point selection strategy, adopts the redundant point elimination strategy and improves the evaluation function. These measures together improve the efficiency of path search. Compared with the traditional A algorithm, the improved algorithm has reduced path length, calculation time, number of traversed nodes, and number of turning points on the path. In particular, the significant reduction in the number of traversed nodes helps to improve the overall computational efficiency.
[0095] To verify the effectiveness of the fusion algorithm, a 25×25 grid map is used. Fig.10 As shown in the figure, known obstacles are black grids, static obstacles are gray, and moving obstacles are yellow squares. In this experiment, the conflict radius D is set to 0.5m to maintain a safe distance between the unmanned logistics vehicle and obstacles. The specific experimental parameters are shown in the following table.
[0096] Table 2 Parameters of unmanned logistics vehicle
[0097]
[0098] Fig.10The simulation results of the improved A fusion DWA algorithm proposed in the present invention on a 25×25 grid map are shown, where the dotted line represents the global guidance path, and the solid line represents the final planned path. It can be seen from the figure that after using the fusion algorithm, the unmanned logistics vehicle can smoothly reach the target point from the starting point, effectively avoiding known obstacles, unknown obstacles and dynamic obstacles, and the path is very smooth, meeting the obstacle avoidance requirements of the unmanned logistics vehicle. Therefore, the unmanned logistics vehicle path planning algorithm based on the fusion of the improved A* algorithm and DWA proposed in the present invention realizes the functions of static global obstacle avoidance and dynamic local obstacle avoidance, showing good effectiveness and feasibility.
[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some or all of the technical features can be replaced by equivalents. However, these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A path planning method for unmanned logistics vehicles based on improved A* and improved DWA fusion algorithms, characterized in that: The following steps are involved: S1. Initialize the parameters and grid map, and set the end point and starting point of the unmanned logistics vehicle path planning; S2. Perform global path planning in the initialized map based on the improved A* algorithm to form a preliminary global path; S3. Process the nodes in the preliminary global path using the tertiary redundant point deletion strategy to generate a global optimized path; S4. Obtain the key turning points in the optimized global path, and use the key turning points as the target points for local path planning. Then, perform local path planning based on the improved DWA algorithm to obtain the optimal motion trajectory.
2. According to claim 1, a path planning method for unmanned logistics vehicles based on the fusion of improved A* and improved DWA is characterized in that ,Based on the modified A* algorithm, global path planning is performed in the initialized map to form a preliminary global path,including: S201, obtaining the starting position, target point position and obstacle information of the unmanned logistics vehicle, and calculating the obstacle density within the rectangle from each subnode to the target point based on the obtained starting position, target point position and obstacle information of the unmanned logistics vehicle; S202, starting from the beginning. Use the evaluation function of the improved algorithm to select the best child node in the neighborhood around the parent node. Then use the best node as the parent node to continue searching for subsequent child nodes until the target position is reached. The weight parameter of the evaluation function of the improved A* algorithm is adjusted by a method of dynamically adjusting the weight of the heuristic function estimation value; S203, backtracking the optimal node extracted in each search to form a preliminary global path.
3. According to claim 2, a path planning method for unmanned logistics vehicles based on the fusion of improved A* and improved DWA is characterized in that ,The nodes in the preliminary global path are processed using the triple redundant point removal method to generate a global optimized path, ,including; S301, the system checks all nodes in the path in sequence, and by determining that the current node is in the same straight line with the two nodes before and after it, the node is considered redundant and deleted from the path nodes; S302, after performing the elimination operation, check the nodes in the path one by one, define the nodes on the path as a set of sequences {Tk|k=1,23…,n}, first find the middle connection point P between T2T3 nodes from T1 node, connect T1 to the middle point P of T2T3, and check whether the path crosses the obstacle area. If yes, consider connecting T1 directly to P, if not, find whether there is a point in the middle of T3T4 that can meet the requirements, until a point is found to connect T1, and continue to infer from this node to find the final point TK; S303. In this case, a new node is obtained, and the previous step is performed again until the final point TK is found, and all nodes after the redundant points are deleted again are obtained.
4. The obstacle avoidance method for unmanned logistics vehicles based on the fusion of improved A* and improved DWA according to claim 3 is characterized in that , obtain the key turning points in the global optimization path, and use the key turning points as local target points, perform local path planning based on the modified DWA algorithm, and obtain the optimal motion trajectory, including; S401. A kinematic model is established based on the motion of the unmanned logistics vehicle. The model uses linear velocity and angular velocity to calculate the position and posture changes of the smart vehicle in continuous time intervals, in which the heading angle is also taken into consideration. The data obtained through sampling accurately describes the small displacement of the smart vehicle along the straight line, and then simulates countless different drivable trajectory routes; S402, using the obtained key turning point as the local target point of the DWA optimization algorithm, using the improved evaluation function to evaluate the quality of the trajectory, and selecting the optimal trajectory; S403, introducing a dynamic safety distance of an obstacle, and evaluating the collision risk according to the moving speed of the obstacle. The faster the obstacle moves, the longer the safety distance is, and the slower the obstacle moves, the shorter the safety distance is. S404, saving the optimal travel trajectory speed space after real-time planning; S405: The unmanned logistics vehicle moves according to the processed trajectory and reaches the set local target point, and immediately determines whether it has reached the final target point. If it has not reached the final point, it returns to step S401 and continues the next stage of path planning until the entire navigation task is completed.
Citation Information
Cited By
Method for managing in-warehouse and out-warehouse data of protective clothing warehouse
CN119313005A
Novel speed evaluation space forklift local path planning method based on DWA
CN119779302A
Unmanned aerial vehicle autonomous flight strategy based on fusion of improved A* algorithm and DWA algorithm
CN119828721A
Local target point selection method and device for rolling route planning of unmanned aerial vehicle
CN120010502A
Trajectory planning method and device for multiple unmanned devices, and unmanned device
CN120029347A
