Mobile robot path planning method based on improved ant colony algorithm

Abstract

The invention belongs to the field of robot path planning, and aims to solve the problem of how to realize better path planning in a complex environment. The invention provides a mobile robot path planning method based on an improved ant colony algorithm, and the method comprises the steps: firstly building environment grid map information, determining a moving starting point and a moving terminal point of each mobile robot, then planning a global path for each mobile robot through employing the improved ant colony algorithm, and carrying out the targeted secondary optimization, therefore, the finally output path is simpler and smoother, the path is shorter, and an effective solution is provided for multi-robot path planning in a complex environment.

Description

Technical Field

[0001] The present invention belongs to the field of robot path planning, and particularly relates to a mobile robot path planning method based on an improved ant colony algorithm. Background Art

[0002] The mobile robot path planning algorithm is an important field in robot technology, which involves how to make the robot autonomously move from one position to another in the environment while avoiding collisions with obstacles during the movement.

[0003] It has broad application prospects in the fields of autonomous driving, industrial automation, service robots, etc. The research and development of mobile robot path planning algorithms are crucial for improving the autonomy, flexibility, and efficiency of robots. These algorithms usually need to consider various factors, such as the dynamic changes of the environment, the dynamic characteristics of the robot, the priority of tasks, and energy consumption, etc.

[0004] Traditional mobile robot path planning algorithms have good effects in the application of simple path planning and are widely used in simple mobile robot path planning problems such as single drones and single unmanned vehicles. However, in the face of complex environments, traditional path planning algorithms are difficult to achieve ideal results. Therefore, applying heuristic intelligent algorithms with learning abilities to path planning is a current trend. Most intelligent optimization algorithms are swarm intelligence algorithms summarized according to certain laws in natural group behaviors and proposed based on these laws. Such algorithms have good solving abilities for high-dimensional complex and multi-constrained optimization problems. Common intelligent optimization algorithms include ant colony algorithm, particle swarm algorithm, genetic algorithm, fish school algorithm, etc.

[0005] The ant colony algorithm is an intelligent bionic algorithm proposed by simulating the foraging behavior of ant colonies in nature. Its core idea is to continuously update the pheromone concentration on the path through positive feedback until the pheromone concentration on the optimal path is the highest, and the optimal solution is found. Due to its characteristics such as positive feedback, parallel computing, and good robustness, the ant colony algorithm has achieved good results in the application of mobile robot path planning. However, the classical ant colony algorithm still has many deficiencies. Therefore, improving the classical ant colony algorithm is particularly important for ensuring the efficient completion of mobile robot path planning. Summary of the Invention

[0006] In order to solve at least one of the above technical problems existing in the prior art, the present invention provides a mobile robot path planning method based on an improved ant colony algorithm.

[0007] The present invention is implemented by adopting the following technical solutions: A mobile robot path planning method based on an improved ant colony algorithm, comprising the following steps: S1: Establish the environmental grid map information, determine the starting and ending points of each mobile robot, and control each mobile robot to move from the starting point according to the environmental grid map information; S2: Randomly select candidate nodes for each mobile robot after the starting point according to the roulette method, and use the improved potential field heuristic function to determine the corresponding optimal moving nodes. Control each mobile robot to move according to the optimal moving nodes until the corresponding mobile robot reaches the end point or cannot move forward. At the same time, judge whether all mobile robots have reached the end point. When there are mobile robots that have not reached the end point, re-execute step S2 until all mobile robots have reached the end point; S3: According to the candidate nodes and the optimal moving nodes, obtain the moving paths of the optimal mobile robot and the worst mobile robot in step S3, as well as the moving paths of the remaining mobile robots, and record the moving path lengths of all mobile robots after the first iteration, and then perform global pheromone update; S4: End the search and judge whether the iteration has reached the maximum number of iterations. If the maximum number of iterations is reached, output the optimal path. If the maximum number of iterations is not reached, return to step S2 and re-execute step S2 and subsequent steps until the maximum number of iterations is reached.

[0008] Preferably, the improved potential field heuristic function in step S2 is: In the formula, is the current moving node to the candidate node the Euclidean distance; is the candidate node the reciprocal of the distance to the nearest obstacle, calculated as: , where is the candidate node the distance to the nearest obstacle; is the safety threshold; is the connectivity evaluation of the candidate node based on the number of all movable next moving nodes around; is the iteration weight term; are the first dynamic weight coefficient, the second dynamic weight coefficient, the third dynamic weight coefficient, and the fourth dynamic weight coefficient in turn, and are dynamically adjusted with the current iteration number dynamically.

[0009] Preferably, step S3 also includes: Obtain the pheromone concentration of the moving paths corresponding to all mobile robots; Update the pheromone concentration using the pheromone concentration update formula.

[0010] Preferably, the pheromone update formula is as follows: In the formula, represents the pheromone concentration of the mobile robot at the -th iteration moving from the mobile node to the candidate node ; is the dynamic global information evaporation factor; is the pheromone concentration of the optimal mobile robot at the -th iteration moving from the mobile node to the candidate node ; is the pheromone concentration of the worst mobile robot at the -th iteration moving from the mobile node to the candidate node ; is the pheromone concentration of the remaining mobile robots at the -th iteration moving from the mobile node to the candidate node ; is the length of the path traveled by the mobile robot at the -th iteration; is the pheromone intensity, which is a constant; is the number of mobile robots that find the shortest path in the current iteration; is the number of mobile robots that find the longest path in the current iteration; represents the length of the shortest path in the current iteration; represents the length of the longest path in the current iteration.

[0011] Preferably, update the dynamic global information evaporation factor , and perform dynamic adjustment according to the number of iterations. The specific formula is as follows: In the formula, is the current number of iterations, is the maximum number of iterations, represents the base of the natural logarithm.

[0012] Preferably, step S3 further includes: Determine the initial movement path according to the movement nodes of all mobile robots, and determine all connected adjacent points on each initial movement path; Remove redundant turning points according to all adjacent points on each initial movement path, and determine the movement path of the optimal mobile robot, the movement path of the worst mobile robot, and the movement paths of the remaining mobile robots according to the initial movement path after removing the turning points.

[0013] Preferably, it further includes: Obtain the rectangular obstacle information in the environmental grid map information; Perform intersection detection based on the initial movement path and the rectangular obstacle information. When the initial movement path intersects with the rectangular obstacle information, the initial movement path is infeasible. When the initial movement path does not intersect with the rectangular obstacle information, the initial movement path is valid.

[0014] Compared with the prior art, the beneficial effects of the present invention are: The present invention provides a mobile robot path planning method based on an improved ant colony algorithm. First, an environmental grid map information is established, and the movement start point and end point of each mobile robot are determined. Then, the improved ant colony algorithm is used to plan a global path for each mobile robot, and targeted secondary optimization is performed, so that the finally output path is simpler and smoother, and the path is shorter, providing an effective solution for multi-robot path planning in a complex environment. Description of the Drawings

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 It is a schematic flowchart of a mobile robot path planning method based on an improved ant colony algorithm provided by an embodiment of the present invention; Figure 2 It is a path planning schematic diagram of a traditional ant colony algorithm provided by an embodiment of the present invention; Figure 3 It is a path planning schematic diagram of an improved ant colony algorithm provided by an embodiment of the present invention; Figure 4 It is a path planning length comparison diagram of a traditional algorithm and an improved ant colony algorithm provided by an embodiment of the present invention; Figure 5 It is a running time comparison diagram of a traditional algorithm and an improved ant colony algorithm provided by an embodiment of the present invention; Figure 6 It is an iteration comparison diagram of a traditional algorithm and an improved ant colony algorithm provided by an embodiment of the present invention. Detailed implementation mode

[0017] Combined with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation manners obtained by those of ordinary skill in the art without creative efforts belong to the scope protected by the present invention.

[0018] It should be noted that the structures, ratios, sizes, etc. shown in the drawings of this specification are only used to cooperate with the content disclosed in the specification for those who are familiar with this technology to understand and read, and are not used to limit the limited conditions under which the present invention can be implemented. Therefore, they do not have technical essence. Any modification of the structure, change of the proportional relationship or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should fall within the scope covered by the technical content disclosed by the present invention. It should be noted that in this specification, relational terms such as first and second are only used to distinguish one entity from several other entities, and do not necessarily require or imply any actual relationship or order between these entities.

[0019] Mobile robots are increasingly widely used in human life, but traditional path planning methods are difficult to meet the path planning of mobile robots in higher - requirement application scenarios. Therefore, to solve the problem of how to achieve better path planning in complex environments, the present invention provides a path planning method for mobile robots based on an improved ant colony algorithm. In this method, the mobile robot is more intelligent during the path search process, the searched path is smoother, more concise, shorter, and the mobile robot avoids collisions with obstacles during the path planning process, ensuring that the generated wheel path is effective and safe.

[0020] To more clearly introduce the above - mentioned objects, features, and advantages of the present invention, the following further detailed description will be made in combination with the accompanying drawings and specific implementation manners.

[0021] As Figure 1 shown, the embodiments of the present invention provide a path planning method for mobile robots based on an improved ant colony algorithm, including the following steps: S1: Establish environmental grid map information, determine the starting point and ending point of each mobile robot's movement, and control each mobile robot to start moving from the starting point according to the environmental grid map information.

[0022] In this embodiment, a suitable lidar camera is selected according to the environmental characteristics of the mobile robot to scan information such as fixed obstacles, unknown obstacles, and moving obstacles in the environment where the mobile robot is located, and at the same time, the positioning of the mobile robot is realized.

[0023] In this embodiment, the ROS (Robot Operating System) software environment suitable for mobile robots is installed and configured. The mobile robot is controlled by a keyboard remote control to move in an indoor environment. While the mobile robot is moving, a lidar is used to scan the environment for simultaneous localization and mapping, and an environmental grid map information and the corresponding coordinate system are established.

[0024] In this embodiment, parameter initialization is performed, including the pheromone weight factor , the heuristic information weight factor , the pheromone evaporation coefficient , the maximum number of iterations , the pheromone intensity , etc., and the grid numbers of the starting point and the end point of the mobile robot are determined.

[0025] S2: After randomly selecting a candidate node after the starting point for each mobile robot according to the roulette method, the improved potential field heuristic function is used to determine the corresponding optimal moving node, and each mobile robot is controlled to move according to the optimal moving node until the corresponding mobile robot reaches the end point, or cannot move forward. At the same time, it is judged whether all mobile robots have reached the end point. When there are mobile robots that have not reached the end point, step S2 is executed again until all mobile robots have reached the end point.

[0026] In this embodiment, at time , the mobile robot is located at the moving node , and it transfers to the candidate node with the largest product of the pheromone concentration and the heuristic information with a standard probability , and takes it as the next moving node. The mobile robot moves on the grid map established according to the environmental grid map information according to this moving rule, and selects the next moving node according to the pseudo-random proportion principle, so as to realize the search and planning of the moving path. The formula of this moving rule is as follows: In the formula, represents the candidate node; represents selecting the candidate node that maximizes from all movable moving nodes, is the set of all movable moving nodes; represents the pheromone concentration from the moving node to the candidate node at time under the influence of the pheromone weight factor , is the pheromone weight factor, which controls the influence degree of pheromone; represents the heuristic information weight factor Under the influence of the moving node at time to the candidate node of the heuristic potential function is the heuristic information weight factor, which controls the influence degree of heuristic information; represents the probability of the neighborhood candidate node at the th iteration;

[0027] When the real-time probability is greater than the standard probability, the next moving node is selected according to the roulette method. The mobile robot calculates the probability of each neighborhood candidate node at the th iteration In the formula, represents the set of neighborhood candidate movable grid cells. In this formula the larger it is, the more inclined the mobile robot is to select the moving path with a large pheromone concentration; the larger it is, the more inclined the mobile robot is to select the node closer to the target point.

[0028] In this embodiment, at any time, the mobile robot moves from the current moving node to the next moving node and then will update in real time the pheromone concentration on the path from the moving node to the next moving node

[0029] In this embodiment, the improved potential field heuristic function is: In the formula, is the Euclidean distance from the current moving node to the candidate node ; is the reciprocal of the distance from the candidate node to the nearest obstacle, calculated as: where is the distance from the candidate node to the nearest obstacle; is the safety threshold; is the connectivity evaluation of the candidate node based on the number of all movable next moving nodes around; is the iteration weight term; They are the first dynamic weight coefficient, the second dynamic weight coefficient, the third dynamic weight coefficient, and the fourth dynamic weight coefficient, which are dynamically adjusted as the current iteration number changes.

[0030] Among them, ; , is the current iteration number, is the maximum iteration number; changes with the iteration number , , , , , is the preset initial weight value, represents the initial maximum weight of the distance term, controlling the weight of the path length in the initial stage. The larger the value, the more inclined the mobile robot is to choose a path with a short distance; represents the initial weight range of the obstacle avoidance term, represents the initial minimum weight of the obstacle avoidance term, ensuring that obstacles will not be completely ignored even in the later stage, represents the initial maximum weight of the obstacle avoidance term, avoiding frequent collisions in the initial stage; represents the initial weight of the connectivity term, controlling the initial weight of the node connectivity evaluation. The larger the value, the more inclined the mobile robot is to choose a path with more feasible nodes around; is the maximum value of the initial iteration weight term, controlling the balance between global exploration (in the initial stage) and local optimization (in the later stage). The larger the value, the more exploration is encouraged in the initial stage; is the decay rate parameter, controlling 's exponential decay speed. The larger the value, decreases faster; controls 's exponential growth speed. The larger the value, increases faster.

[0031] In this embodiment, in order to make the heuristic potential function better adapt to complex environments and dynamic adjustment requirements in the ant colony algorithm, a potential field heuristic function including multiple node information, the current iteration number and the maximum iteration number is designed. By guiding the path away from obstacles, avoiding the selection of isolated nodes, and improving the path feasibility. During the initial stage of dynamic iterative adjustment , high and emphasize the distance to the target and global exploration, while low and relax the obstacle avoidance requirements; during the later stage of dynamic iterative adjustment When it is low and reduce randomness, high and strictly avoid obstacles and optimize path quality; finally, balance exploration and exploitation, and gradually shift from global exploration to local optimization to avoid falling into local optima by gradually shifting from global exploration to local optimization to avoid falling into local optima.

[0032] The mobile robot starts path search and uses a loop to simulate the search process of each mobile robot. A total of rounds of iterative search are performed. In each round of search, each mobile robot starts from the starting point of movement, randomly selects a movement node after the starting point of movement according to the roulette method at any time, and determines the corresponding optimal movement node with an improved heuristic function. Then the mobile robot moves according to the selected optimal movement node until the mobile robot reaches the end point or cannot move forward.

[0033] S3: According to the movement node and the optimal movement node, obtain the movement path of the optimal mobile robot, the movement path of the worst mobile robot, and the movement paths of the remaining mobile robots in step S3, and record the movement path lengths of all mobile robots after the first iteration, and then perform global pheromone update.

[0034] Optionally, it further includes: obtaining the pheromone concentration of the movement paths corresponding to all mobile robots; updating the pheromone concentration by using the pheromone concentration update formula.

[0035] In this embodiment, through the pheromone reward and punishment strategy, the pheromone of movement paths with different qualities is updated, so as to achieve better effects in both global search and local optimization.

[0036] First, according to the length of the movement path passed by the mobile robot, the corresponding mobile robot is defined as the following types: Optimal mobile robot: For those mobile robots that have found the shortest path, the movement paths they pass through are considered high-quality paths, so the corresponding mobile robots are called optimal mobile robots. When updating the pheromone, a larger pheromone increment is given to the optimal mobile robots, so as to enhance the influence of these paths. Denote the iteration times of the optimal mobile robot from the movement node to the candidate node The pheromone concentration is defined as: In the formula, is the dynamic global information evaporation factor; Denote the iteration times of the optimal mobile robot from the movement node to the candidate node pheromone concentration; is the th iteration of the mobile robot the length of the path traveled; is the pheromone intensity, which is a constant; is the number of mobile robots that find the shortest path in the current iteration; represents the shortest path length of the current iteration; Worst mobile robots: For those mobile robots that find the longest paths, the paths they pass through are considered low-quality paths, so the corresponding mobile robots are called the worst mobile robots. When updating the pheromone, a smaller pheromone increment is given to the worst mobile robots, thus reducing the influence of these paths. represents the th iteration of the worst mobile robot from the mobile node to the candidate node the pheromone concentration is defined as: where is the number of mobile robots that find the longest path in the current iteration; represents the longest path length of the current iteration; represents the th iteration of the worst mobile robot from the mobile node to the candidate node pheromone concentration.

[0037] Other mobile robots: For other mobile robots, the paths they pass through are considered sub-optimal. When updating the pheromone, a regular pheromone increment is given to other mobile robots to maintain the influence of these paths. represents the th iteration of other mobile robots from the mobile node to the candidate node the pheromone concentration is defined as: where represents the th iteration of other mobile robots from the mobile node to the candidate node pheromone concentration.

[0038] At any time, after the mobile robot moves from the current mobile node to the next mobile node, it will update the pheromone concentration on this path in real time. After all mobile robots complete the initial iteration, record the number of the worst mobile robots and the number of the best mobile robots during this iteration process. At the same time, record the moving path lengths of all mobile robots, and update the global pheromone according to the improved global pheromone formula.

[0039] When all mobile robots complete an iterative search, update the pheromone concentration of the moving path of the globally optimal mobile robot, and perform real-time update using the improved pheromone concentration formula.

[0040] The pheromone update formula is as follows: In this embodiment, compared with the global pheromone evaporation factor in the traditional ant colony algorithm is a fixed constant, which leads to unreasonable pheromone distribution when the mobile robot searches for a path. To avoid this defect, change the global pheromone evaporation factor from a static value to a value that is dynamically adjusted according to the number of iterations , and the specific formula is: In the formula, is the current number of iterations, is the maximum number of iterations, represents the base of the natural logarithm.

[0041] By adaptively updating the global pheromone evaporation factor, making it smaller in the initial stage of iteration and gradually increasing, it prompts the mobile robot to explore more; in the later stage of iteration, as the number of iterations increases, the global pheromone evaporation factor increases, causing the pheromone concentration to increase, making the guiding effect of the pheromone concentration on the mobile robot stronger, and then correspondingly reducing the search time of the mobile robot, better balancing the global search and the local search.

[0042] Optionally, it further includes: determining an initial moving path according to the moving nodes of all mobile robots, and determining all connected adjacent points on each initial moving path; removing redundant turning points according to all adjacent points on each initial moving path, and determining the moving path of the best mobile robot, the moving path of the worst mobile robot, and the moving paths of other mobile robots according to the initial moving path after removing the turning points.

[0043] In this embodiment, the traditional ant colony algorithm is an eight-degree-of-freedom search probability algorithm derived from the foraging behavior of ant colonies. The step size selection of the ant colony algorithm is simple, which limits the efficiency of the path planning of mobile robots. To solve this problem, redundant turning points are eliminated through secondary optimization while reducing the path length. In this application, redundant turning points are removed by judging the connectivity of adjacent points on the path. The specific process is as follows: Step 1: Starting point setting and node addition: Starting from the starting point of the initial movement path, the current moving node is taken as a node of the new path and added to the new path; Step 2: Searching for connected adjacent points: Starting from the current moving node, traverse backward to the last moving node of the initial movement path, and gradually check whether the adjacent points in each feasible area are connected to the current moving node. Connectivity can be understood as no obstacle blocking between two points, that is, they can be directly connected. When a connected adjacent point is found, set this adjacent point as the starting point of the new path and proceed to the next step; Step 3: Removing redundant turning points: Once a connected adjacent point is found, take it as the new current point and continue to search for the next connected adjacent point in the direction of the starting point of the initial movement path. This will directly connect the adjacent points and eliminate the redundant turning points in the initial movement path; Step 4: Path validity verification: Before connecting adjacent points, obstacle collision detection needs to be performed. When the straight-line path connecting adjacent points intersects with an obstacle, it means this path is not feasible, and then a connected adjacent point needs to be reselected; Step 5: Repeating the optimization steps: Repeat steps 2 to 4 until reaching the last moving node of the initial movement path, or until no more connected adjacent points can be found.

[0044] Step 6: Outputting the optimized path: After the above steps, the obtained path is the optimized path, where the redundant turning points have been eliminated and the path is smoother and more concise.

[0045] Optionally, it further includes: obtaining the rectangular obstacle information in the environmental grid map information; performing intersection detection based on the initial movement path and the rectangular obstacle information. When the initial movement path intersects with the rectangular obstacle information, the initial movement path is not feasible. When the initial movement path does not intersect with the rectangular obstacle information, the initial movement path is valid.

[0046] In this embodiment, during the path optimization process, obstacle collision detection also needs to be considered. Obstacle collision detection is based on the intersection detection between a line segment and a rectangular obstacle. By detecting whether the line segment on the movement path intersects with an obstacle, it is judged whether the path is valid. If a certain segment of the movement path intersects with an obstacle, it means the initial movement path is not feasible.

[0047] In this embodiment, in obstacle collision detection, assume that the initial movement path and the rectangular obstacle are line segment AB and line segment CD respectively, and the cross product of vectors is calculated to determine whether they intersect. Specifically, assume that the starting point of line segment AB is point A, the ending point is point B, the starting point of line segment CD is point C, and the ending point is point D. The cross product formula is as follows: Wherein, represents the vector corresponding to line segment AB, represents the vector corresponding to line segment CD, and the result of the cross product cross can be a scalar.

[0048] According to the result of the cross product, the intersection situation of the two line segments can be judged: If the cross product cross is positive, it means that line segment AB intersects CD, that is, the initial movement path is not feasible.

[0049] If the cross product cross is zero, it means that line segment AB is parallel to or overlaps with CD, that is, the initial movement path is valid.

[0050] If the cross product cross is negative, it means that line segment AB does not intersect CD, that is, the initial movement path is valid.

[0051] In path optimization, the cross product is used to judge whether the optimized path segment intersects with the obstacle boundary. Specifically, when judging whether the path segment to intersects with the obstacle boundary, the following cross product calculation can be used: In the formula, is the maximum coordinate point of the obstacle boundary. If the cross product cross is positive, it means that the path segment intersects with the obstacle boundary, and this path needs to be avoided. If the cross product cross is zero or negative, it means that the path does not intersect with the obstacle, and the path is feasible.

[0052] In summary, the cross product is used in path planning to judge whether line segments intersect, helping us avoid collisions between the path and obstacles, thereby ensuring that the generated path is effective and safe.

[0053] S4: End the search, and judge whether the iteration reaches the maximum number of iterations. If the maximum number of iterations is reached, output the optimal path. If the maximum number of iterations is not reached, return to step S2 and re - execute step S2 and subsequent steps until the maximum number of iterations is reached.

[0054] In this embodiment, when the maximum number of iterations is not reached, clear the taboo table in the environmental grid map information, and let the mobile robot , return to step S2 and loop through the subsequent steps in sequence until the maximum number of iterations is reached, and output the optimal path length.

[0055] In this embodiment, as Figure 2 and Figure 3 respectively represent the path planning diagram of the mobile robot by the traditional ant colony algorithm and the path planning diagram of the mobile robot output by the embodiment of the present invention. The black square represents a fixed obstacle, the gray cross square is an unknown obstacle, and the light slash square is a moving obstacle. The lengths of the horizontal and vertical coordinate systems are customized according to the actual situation. Through Figure 2 and Figure 3 comparison, it can be seen that after multiple iterations, the optimal path output by the improved ant colony algorithm of the present invention is smoother, more concise, and shorter than that of the traditional ant colony algorithm.

[0056] In this embodiment, as Figure 4 shows the path planning length diagram of the original algorithm and the improved ant colony algorithm provided by this application. The two algorithms are used for experimental tests respectively, with 20 samples each. The path planning length range of the original algorithm is from 27.5 cm to 28.2 cm, and the path planning length range of the improved ant colony algorithm is from 25.3 cm to 26.2 cm. Through comparison, it can be seen that the path planning length of the improved ant colony algorithm has decreased by an average of 2.1 cm compared with the original algorithm.

[0057] As Figure 5 shows the running time diagram of the original algorithm and the improved ant colony algorithm. The two algorithms are used to test 20 samples respectively. The time range of the original algorithm is from 9.24 seconds to 10.25 seconds, and the time range of the improved ant colony algorithm is from 7.25 seconds to 8.26 cm. Through comparison, it can be seen that the running time of the improved ant colony algorithm has decreased by an average of 1.9 seconds compared with the original algorithm.

[0058] As Figure 6 shows the algorithm iteration training diagram of the original algorithm and the improved ant colony algorithm. It can be seen from Figure 6 that the original algorithm has a slow convergence speed and reaches convergence after 78 trainings, with an optimal length of 28.2 cm. The improved ant colony algorithm has a fast convergence speed and reaches convergence after 47 trainings, with an optimal length of 25.5 cm.

[0059] The present invention optimizes the potential field heuristic function, updates the pheromone concentration according to the improved pheromone reward and punishment strategy, and simultaneously realizes the adaptive adjustment of the global evaporation factor, making the mobile robot more intelligent in the path planning process, obtaining a smoother, more concise, and shorter moving path.

[0060] As described above, it is only the preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims described above.

Claims

1. A path planning method for a mobile robot based on an improved ant colony algorithm, characterized in that, It includes the following steps: S1: Establish environmental grid map information, determine the starting and ending points of each mobile robot, and control each mobile robot to move starting from the starting point according to the environmental grid map information; S2: Randomly select candidate nodes after the starting point for each mobile robot according to the roulette method, and use an improved potential field heuristic function to determine the corresponding optimal moving nodes. Control each mobile robot to move according to the optimal moving nodes until the corresponding mobile robot reaches the ending point or cannot move forward. At the same time, judge whether all mobile robots have reached the ending point. When there are mobile robots that have not reached the ending point, re-execute step S2 until all mobile robots have reached the ending point; S3: According to the candidate nodes and the optimal moving nodes, obtain the moving paths of the optimal mobile robot, the worst mobile robot, and the remaining mobile robots in step S3, and record the moving path lengths of all mobile robots after the first iteration, and then perform global pheromone update; S4: End the search and judge whether the iteration has reached the maximum number of iterations. If the maximum number of iterations is reached, output the optimal path. If the maximum number of iterations is not reached, return to step S2 and re-execute step S2 and subsequent steps until the maximum number of iterations is reached.

2. The mobile robot path planning method based on the improved ant colony algorithm according to claim 1, wherein The improved potential field heuristic function in step S2 is: Wherein, is the Euclidean distance from the current moving node to the candidate node ; is the reciprocal of the distance from the candidate node to the nearest obstacle, calculated as: , where is the distance from the candidate node to the nearest obstacle; is the safety threshold; is the connectivity evaluation of the candidate node , based on the number of all movable next moving nodes around; is the iterative weight term; are the first dynamic weight coefficient, the second dynamic weight coefficient, the third dynamic weight coefficient, and the fourth dynamic weight coefficient in sequence, and are dynamically adjusted with the current iteration number .

3. A path planning method for a mobile robot based on an improved ant colony algorithm according to claim 1, characterized in that, Step S3 also includes: Obtain the pheromone concentrations of the moving paths corresponding to all mobile robots; Update the pheromone concentrations using the pheromone concentration update formula.

4. A path planning method for a mobile robot based on an improved ant colony algorithm according to claim 3, wherein, The pheromone update formula is: Wherein, represents the pheromone concentration of the mobile robot after iterations from the mobile node to the candidate node ; is the dynamic global information volatile pheromone; is the pheromone concentration of the optimal mobile robot from the mobile node to the candidate node at the th iteration; is the pheromone concentration of the worst mobile robot from the mobile node to the candidate node at the th iteration; is the pheromone concentration of the remaining mobile robots from the mobile node to the candidate node at the th iteration; is the length of the path traveled by the mobile robot at the th iteration; is the pheromone intensity, which is a constant; is the number of mobile robots that find the shortest path in the current iteration; is the number of mobile robots that find the longest path in the current iteration; represents the shortest path length of the current iteration; represents the longest path length of the current iteration.

5. A path planning method for a mobile robot based on an improved ant colony algorithm according to claim 4, characterized in that, Update the volatile element of the dynamic global information , and perform dynamic adjustment according to the number of iterations. The specific formula is as follows: In the formula, is the current iteration number, is the maximum iteration number, represents the base of the natural logarithm.

6. A path planning method for a mobile robot based on an improved ant colony algorithm according to claim 1, characterized in that, Step S3 also includes: Determine the initial moving paths according to the moving nodes of all mobile robots, and determine all connected adjacent points on each initial moving path; Remove redundant turning points according to all adjacent points on each initial moving path, and determine the moving paths of the optimal mobile robot, the worst mobile robot, and the remaining mobile robots according to the initial moving paths after removing the turning points.

7. A path planning method for a mobile robot based on an improved ant colony algorithm according to claim 6, characterized in that, It also includes: Obtain the rectangular obstacle information in the environmental grid map information; Perform intersection detection based on the initial moving paths and the rectangular obstacle information. When the initial moving path intersects with the rectangular obstacle information, the initial moving path is infeasible. When the initial moving path does not intersect with the rectangular obstacle information, the initial moving path is valid.

Images