Automatic guided vehicle path planning method and system based on pheromone limitation updating
The ant colony algorithm, which uses pheromone-restricted updates and direction factor optimization, solves the problems of slow convergence speed and easy getting trapped in local optima in the path planning of automated guided vehicles, and achieves faster path search and higher stability.
Patent Information
- Application Number
- CN202511193086.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-25
- Publication Date
- 2025-11-21
AI Technical Summary
Existing automated guided vehicle (AGV) path planning algorithms have slow convergence speed in the initial stage, are prone to getting trapped in local optima, lack dynamic environment adaptability, and have uneven paths.
The ant colony algorithm with pheromone-constrained updates optimizes the path planning process by enhancing pheromone through local optimal paths and introducing a direction factor, thereby avoiding indiscriminate updates globally and improving the stability and goal orientation of path search.
It accelerates the convergence speed of the ant colony algorithm, reduces the number of iterations and path length, improves the efficiency and reliability of path search, and adapts to dynamic environments.
Smart Images
Figure CN120993914A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of path planning technology, specifically relating to an automated guided vehicle path planning method and system based on pheromone-constrained updates. Background Technology
[0002] AGV (Automated Guided Vehicle) is an unmanned intelligent transportation device based on electromagnetic, optical, or laser navigation technology, capable of precisely traveling along a preset path and completing material handling tasks. As a prime example of the industrial application of wheeled mobile robots, the AGV system has three core characteristics: powered by batteries, equipped with a non-contact navigation device, and integrated intelligent safety protection system.
[0003] Compared to manual sorting and handling, automated guided vehicles (AGVs) solutions significantly improve operational accuracy (±10mm positioning) and efficiency (24 / 7 continuous operation). Their applications cover high-end fields such as automotive manufacturing, precision electronics machining, and pharmaceutical cold chain logistics, and they demonstrate superior flexibility in scenarios such as port automation and smart warehousing. Existing AGVs typically employ path planning algorithms such as A*, RRT, ant colony optimization, and genetic algorithms to generate paths from the initial point to the target point.
[0004] Traditional ant colony optimization (ACO) algorithms simulate ant pheromone behavior to solve path optimization problems. Their advantages in robot path planning include distributed characteristics and adaptability. However, existing algorithms suffer from several problems: slow convergence in the initial stage, susceptibility to local optima, lack of dynamic environmental adaptability, and uneven path smoothness. Summary of the Invention
[0005] The purpose of this invention is to provide an automated guided vehicle (AGV) path planning method and system based on pheromone-restricted updates.
[0006] In a first aspect, the present invention provides an automated guided vehicle (AGV) path planning method based on pheromone-constrained updates, the method comprising:
[0007] The system acquires information on the starting point, ending point, and obstacles of the automated guided vehicle (AGV), and generates a two-dimensional grid map based on the acquired data. On the generated two-dimensional grid map, the ant colony algorithm is used to perform global path planning search to obtain the optimal path for the AGV from the starting point to the ending point.
[0008] In the process of global path planning search using the ant colony algorithm, the locally optimal path is obtained in each iteration, and after attenuating the pheromone on all paths, the pheromone on the locally optimal path is enhanced only.
[0009] Preferably, the pheromone enhanced by the local optimal path is the sum of the attenuated pheromone and the enhancement amount; the enhancement amount is inversely proportional to the path length of the local path.
[0010] Preferably, during the global path planning search process, the target transfer node for the ant is selected from all candidate transfer nodes by introducing the transfer probability of a direction factor; the direction factor is obtained based on the vector formed by a single candidate transfer node and the current node; the transfer probability... The method to obtain it is as follows:
[0011]
[0012] Where, τ ij and The pheromone intensity between different nodes; η ij and For heuristic functions; D ij and α is the directional factor between different nodes; β is the set of candidate transition nodes; γ is the pheromone importance factor; β is the heuristic function importance factor; and γ is the factor that adjusts the directional guidance strength.
[0013] Preferably, the method for obtaining the direction factor is as follows:
[0014]
[0015] in, Direction factor; The direction indicator parameter is expressed as follows:
[0016]
[0017] in, This is the vector from the current node to the endpoint; Let be the vector from the current node to the candidate transition node j.
[0018] Preferably, the candidate transfer node is a node surrounding the current node that has not been set as the target transfer node.
[0019] Preferably, the heuristic function is the reciprocal of the Euclidean distance between different nodes plus 1.
[0020] Preferably, the specific process of using the ant colony algorithm for global path planning and search is as follows:
[0021] Initialize ant colony parameters; each ant starts its search from the starting point and selects a target transfer node based on the transfer probability until the target transfer node becomes the destination, thus completing the ant's search; after all ants have completed their search, obtain the local optimal path by comparing the paths of different ants; obtain the global optimal path based on the local optimal paths in different iterations, and use the global optimal path when the iteration termination condition is met as the optimal path for the automated guided vehicle from the starting point to the destination.
[0022] Secondly, the present invention provides an automated guided vehicle (AGV) path planning system based on pheromone-restricted updates, which is used to execute the aforementioned AGV path planning method. The AGV path planning system includes a data acquisition module, a raster map generation module, a node update module, and a path generation module. The data acquisition module is used to collect environmental information and input it into the raster map generation module to generate a raster map. The node update module selects transfer nodes by constructing transfer probabilities and inputs the selected transfer nodes into the path generation module to obtain the optimal path.
[0023] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor; the memory stores the computer program; and the processor executes the above-described automated guided vehicle path planning method.
[0024] Fourthly, the present invention provides a readable storage medium storing a computer program; when executed by a processor, the computer program is used to implement the above-described automated guided vehicle path planning method.
[0025] The beneficial effects of this invention are:
[0026] 1. This invention avoids the slow convergence speed caused by indiscriminate updates in the global scope by updating only the pheromone on the locally optimal path in each iteration, thereby effectively accelerating the convergence speed of the ant colony algorithm to obtain the globally optimal path.
[0027] 2. This invention introduces a direction factor based on the movement direction in the construction of the transition probability, which enables ants to have stronger goal orientation during the path search process, and can move towards the destination direction with a higher probability. This effectively reduces invalid searches and detours, avoids getting trapped in local optima, improves the global optimization ability of path search, and enhances the stability and reliability of path search. Attached Figure Description
[0028] Figure 1 This is the overall flowchart of the present invention.
[0029] Figure 2 This is a simulation path result diagram of the traditional ant colony algorithm on a simple grid map.
[0030] Figure 3 This is a graph showing the number of simulation path iterations for the traditional ant colony algorithm on a simple grid map.
[0031] Figure 4 This is a simulation path result diagram of the present invention on a simple grid map.
[0032] Figure 5 This is a diagram showing the number of simulation path iterations for this invention on a simple grid map.
[0033] Figure 6 This is a simulation path result diagram of the traditional ant colony algorithm on a complex grid map.
[0034] Figure 7 This is a graph showing the number of simulation path iterations for the traditional ant colony algorithm on a complex grid map.
[0035] Figure 8 This is a simulation path result diagram of the present invention under a complex grid map.
[0036] Figure 9 This is a diagram showing the number of simulation path iterations for this invention on a complex grid map. Detailed Implementation
[0037] The present invention will be further described below with reference to the accompanying drawings.
[0038] An automated guided vehicle (AGV) path planning method based on pheromone-constrained updates is proposed. The AGV path planning system includes a data acquisition module, a raster map generation module, a node update module, and a path generation module. The data acquisition module collects environmental information using a lidar or depth camera mounted on the AGV and inputs it into the raster map generation module to generate a raster map. The node update module selects transition nodes by constructing transition probabilities and inputs the selected transition nodes at different times into the path generation module to obtain the optimal path.
[0039] like Figure 1 As shown, the automated guided vehicle (AGV) path planning method includes the following steps:
[0040] Step 1: The LiDAR or depth camera mounted on the Automated Guided Vehicle (AGV) collects real-time information on the AGV's starting and ending points, as well as obstacles. SLAM (Simultaneous Localization and Mapping) is then used to generate a two-dimensional grid map based on the collected data. Each cell on the two-dimensional grid map represents an obstacle or a passable area.
[0041] Step 2: Initialize the ant colony size (number of ants) M, pheromone importance factor α, heuristic function importance factor β, pheromone evaporation factor ρ, pheromone constant Q, and the maximum number of iterations itermax. The number of ants is the number of individual ants simultaneously searching a path; the pheromone importance factor controls the degree of ant dependence on pheromones; the heuristic function importance factor measures the guiding strength of the target direction on ant behavior; the pheromone evaporation factor simulates the natural decay of pheromones over time; the pheromone increment constant affects the intensity of pheromone updates; and the maximum number of iterations controls the total number of iterations required for algorithm convergence.
[0042] Step 3: Select the target transfer node
[0043] Since existing ant colony algorithms only consider pheromone concentration and heuristic functions in their transition probabilities, this paper introduces a direction guidance mechanism into the transition probability. The ant's movement direction is used as an important reference, and the target transition node is selected from all candidate transition nodes based on the transition probability. The expression is:
[0044]
[0045] Where, τ ij and The pheromone intensity between different nodes; η ij and D is a heuristic function representing the degree of expectation between different nodes. ij and α is the directional factor between different nodes; β is the set of candidate transition nodes; γ is the pheromone importance factor; β is the heuristic function importance factor; and γ is the factor that adjusts the directional guidance strength.
[0046] The larger the pheromone importance factor α, the greater the probability that ants will choose previously visited paths, reducing the randomness of the search path and decreasing the search range of the ant colony, making it prone to getting trapped in local optima. The larger the heuristic function importance factor β, the greater the role of the heuristic function in the transition, meaning ants will move to shorter nodes with a higher probability. The ant colony is more likely to choose shorter local paths, accelerating the algorithm's convergence speed, but reducing randomness and making it easier to reach local optima. Path movement follows the "finite state transition probability rule": ants do not backtrack and avoid visiting already visited nodes; that is, they use the surrounding nodes of the ant's current node that are not set as the target transition node as candidate transition nodes.
[0047] Using the Euclidean distance formula as the heuristic function for the ant colony algorithm, its expression is:
[0048]
[0049] Where, d ij This represents the Euclidean distance between different nodes.
[0050] Initially, the set allowed contains (n-1) elements, which includes all nodes except the starting node of the ant. As time progresses, the number of elements in the set allowed decreases until it becomes empty, indicating that all nodes have been visited.
[0051] Direction factor D ij Based on the ant's movement direction (the direction from its current position to the transfer node), its expression is:
[0052]
[0053] in, The direction indicator parameter is expressed as follows:
[0054]
[0055] in, Let be the vector from node i to the endpoint; Let be the vector from node i to node j.
[0056] If direction indicator parameters A value of 1 (optimal solution) indicates that the ant moves entirely in the direction of the destination; if the direction indicator parameter is 1... A value of 0 (medium solution) indicates that the ant's movement direction is perpendicular to the direction from its current position to the destination; if the direction indicator parameter... A value of -1 (worst-case solution) indicates that the ant's movement direction is opposite to the direction from its current position to the destination.
[0057] By using the direction factor D ij Mapped to the [0,1] interval for easy computation.
[0058] Step 4: Obtain the local optimal path
[0059] Repeat step three to select the target transfer node until the target transfer node is the endpoint, and obtain the current ant's path based on all target transfer nodes. Compare the paths of all ants in the ant colony and select the locally optimal path in the current iteration. Compare the locally optimal path with the globally optimal path. If the locally optimal path is better than the globally optimal path, then the locally optimal path is taken as the globally optimal path.
[0060] In this embodiment, the locally optimal path in the initial iteration process is selected as the globally optimal path.
[0061] Step 5: Pheromones Update
[0062] In the traditional ant colony algorithm's pathfinding process, each ant starts from the starting point and releases a certain amount of pheromone for each unit it moves (in any of the four connected directions: up, down, left, and right). If the path is short, the accumulated pheromone concentration will be high, attracting more ants to choose it in subsequent iterations, thus achieving a "positive feedback mechanism."
[0063] 5-1. Natural volatilization mechanism
[0064] All pheromones from all pathways are allowed to decay naturally in proportion to (1-ρ) to prevent an excessively strong "path memory effect." The decayed pheromones... The expression is:
[0065]
[0066] Where ρ is the pheromone evaporation factor.
[0067] 5-2. Optimal Path Enhancement Mechanism
[0068] Enhance the pheromone on the locally optimal path, with the enhancement amount inversely proportional to the path length. The enhanced pheromone... The expression is:
[0069]
[0070] in, The pheromone constant; This is a locally optimal path.
[0071] Step 6: Repeat steps 3 to 5 until the current iteration count reaches the maximum iteration count itermax or all ants converge to the same path. Use the globally optimal path as the optimal path for the automated guided vehicle from the starting point to the end point.
[0072] Step 7: Conduct algorithm simulation experiments using MATLAB software to verify the feasibility of the algorithm. The simulation environments are a simple raster map and a complex raster map. The experimental results obtained in the 20*20 simple raster map environment are as follows: Figure 2 , 3 4, 5 and Table 1 are shown.
[0073] Table 1. Experimental results in a 20*20 simple raster map environment.
[0074] algorithm Path length Search time / ms Number of iterations / times Traditional ant colony algorithm 23.9706 709.31 17 This invention 22.5563 532.75 4
[0075] As can be seen from Table 1, in a simple environment, the number of iterations of the present invention is reduced by more than 76%, the path length is shortened by more than 5.9%, and the search time is also faster than that of the traditional ant colony algorithm.
[0076] The experimental results obtained in a 40*40 complex raster map environment are as follows: Figure 6 , 7 8, 9 and Table 2 are shown.
[0077] Table 2. Experimental results under a 40*40 complex raster map environment.
[0078] algorithm Path length Search time / ms Number of iterations / times Traditional ant colony algorithm 72.4264 1278.2 82 This invention 46.5269 1093.6 19
[0079] As can be seen from Table 2, in complex environments, the number of iterations of the present invention is reduced by up to 76%, the path length is shortened by more than 35%, and the search time is also faster than that of the traditional ant colony algorithm.
Claims
1. A method for path planning of automated guided vehicles based on pheromone-constrained updates, characterized in that: The method includes: The system acquires information on the starting point, ending point, and obstacles of the automated guided vehicle (AGV), and generates a two-dimensional grid map based on the acquired data. On the generated two-dimensional grid map, the ant colony algorithm is used to perform global path planning search to obtain the optimal path for the AGV from the starting point to the ending point. In the process of global path planning search using the ant colony algorithm, the locally optimal path is obtained in each iteration, and after attenuating the pheromone on all paths, the pheromone on the locally optimal path is enhanced only.
2. The method for automatic guided vehicle path planning based on pheromone-restricted updates according to claim 1, characterized in that: The enhanced pheromone of the local optimal path is the sum of the attenuated pheromone and the enhancement amount; the enhancement amount is inversely proportional to the path length of the local path.
3. The method for automatic guided vehicle path planning based on pheromone-restricted updates according to claim 1, characterized in that: During the global path planning search process, the target transfer node for the ant is selected from all candidate transfer nodes by introducing the transfer probability of the direction factor; The direction factor is obtained based on the vector formed by a single candidate transition node and the current node; the transition probability The method to obtain it is as follows: ； Where, τ ij and The pheromone intensity between different nodes; η ij and For heuristic functions; D ij and α is the directional factor between different nodes; β is the set of candidate transition nodes; γ is the pheromone importance factor; β is the heuristic function importance factor; and γ is the factor that adjusts the directional guidance strength.
4. The method for automatic guided vehicle path planning based on pheromone-restricted updates according to claim 3, characterized in that: The method for obtaining the direction factor is as follows: ； in, Direction factor; The direction indicator parameter is expressed as follows: ； in, This is the vector from the current node to the endpoint; Let be the vector from the current node to the candidate transition node j.
5. The method for automatic guided vehicle path planning based on pheromone-restricted updates according to claim 3, characterized in that: The candidate transfer node is a node surrounding the current node that has not been set as the target transfer node.
6. The method for automatic guided vehicle path planning based on pheromone-restricted updates according to claim 3, characterized in that: The heuristic function is the reciprocal of the Euclidean distance between different nodes plus 1.
7. The method for automatic guided vehicle path planning based on pheromone-restricted updates according to claim 1, characterized in that: The specific process of using the ant colony algorithm for global path planning and search is as follows: Initialize ant colony parameters; each ant starts its search from the starting point and selects a target transfer node based on the transfer probability until the target transfer node becomes the destination, thus completing the ant's search; after all ants have completed their search, obtain the local optimal path by comparing the paths of different ants; obtain the global optimal path based on the local optimal paths in different iterations, and use the global optimal path when the iteration termination condition is met as the optimal path for the automated guided vehicle from the starting point to the destination.
8. An automated guided vehicle (AGV) path planning system based on pheromone-constrained updates, characterized in that: The system is used to execute the pheromone-restricted update-based automated guided vehicle (AGV) path planning method according to claim 1; the AGV path planning system includes a data acquisition module, a grid map generation module, a node update module, and a path generation module; The data acquisition module collects environmental information and inputs it into the raster map generation module to generate a raster map; the node update module selects transfer nodes by constructing transfer probabilities and inputs the selected transfer nodes into the path generation module to obtain the optimal path.
9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The memory stores a computer program; the processor executes an automated guided vehicle path planning method based on pheromone-restricted updates as described in any one of claims 1-7.
10. A readable storage medium storing a computer program; characterized in that: When the computer program is executed by the processor, it is used to implement an automated guided vehicle path planning method based on pheromone-restricted updates as described in any one of claims 1-7.