A method and system for mobile robot path planning

Abstract

The present invention discloses a method and system for mobile robot path planning, including: (1) initializing a map; (2) setting a cost function; (3) initializing the position and speed of each particle; (4) calculating the fitness value of each particle; (5) taking a random number and comparing it with a decision number, if it is greater than the decision number, executing (6), otherwise executing (7); (6) for each particle, updating the individual optimal value, the global optimal value and the speed, position and fitness value, and executing (9); (7) selecting inferior particles from all particles to perform mutation operations to obtain mutant particles; (8) combining the original particle group with the mutant particles to form a new particle group; calculating the fitness value of the new particle group and sorting it, and eliminating the K particles at the end; (9) if the maximum number of iterations is not reached, returning to step (5); if it is reached, selecting the particle with the smallest fitness value as the path solution. The present invention solves the problems of poor path selection diversity and slow convergence speed of the particle swarm algorithm in mobile robot path planning.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a method and system for mobile robot path planning. Background Art

[0002] The path planning of mobile robots is one of the core research areas of mobile robots. Its core is to guide the robot to move from a given starting point to a target point without collision. In this process, the path is optimized, the time is shortest, and the path is smoothest as much as possible. In the prior art, the path planning problem is converted into an optimization problem, and the particle swarm algorithm is introduced into the optimization problem to solve it, and finally the optimal path is obtained. First, the optimization using the particle swarm algorithm depends on the result of algorithm initialization. If the initialization result is not good, it will affect the convergence speed of the algorithm; the fixed parameters of the particle swarm algorithm also have a great influence on the algorithm results. Different parameters will affect whether the algorithm globally expands the search or locally explores in this iteration, affecting the convergence speed of the algorithm; finally, the algorithm does not improve the particles with poor performance in the algorithm, which greatly wastes computing power in each iteration and affects the diversity of the algorithm. Summary of the invention

[0003] Purpose of the invention: In view of the above shortcomings, the present invention provides a method for mobile robot path planning to solve the problems of poor path selection diversity, slow convergence speed and easy to fall into local optimum in the path planning of mobile robots by particle swarm algorithm. At the same time, a system for mobile robot path planning is also provided.

[0004] Technical solution: To solve the above problems, the present invention provides a method for mobile robot path planning, comprising the following steps:

[0005] (1) Formatting the robot's activity area map by using a grid map and identifying obstacle-free areas and obstacle-affected areas; storing the starting point and end point of the robot's movement;

[0006] (2) Set the cost function of the particle swarm algorithm, the formula is as follows:

[0007]

[0008] Where L(p) is the shortest path obtained by the particle, L o is the distance between the mobile robot and the nearest obstacle; α and β are fixed parameters used to control the optimal path that balances path length and safety; F is the required cost function value;

[0009] (3) Initialize the parameters of the particle swarm algorithm; set the number of particles and the total number of iterations of the particle swarm algorithm, set the particle position limit and particle speed limit conditions according to the formatted map, and randomly initialize the position and speed of each particle;

[0010] (4) After obtaining the position and velocity of each particle in the particle swarm, substitute the particle position into the cost function to obtain the fitness value of each particle, and save the fitness value of each particle;

[0011] (5) Take a random number and compare it with the decision number. If the value of the random number is greater than the decision number, execute step (6); otherwise, execute step (7). The value of the random number is [0, 1]. The formula of the decision number is as follows:

[0012]

[0013] In the formula, Q is the number of decisions; t is the current number of iterations, T max The maximum number of iterations set;

[0014] (6) For each particle in the particle swarm, update the individual optimal and global optimal values ​​and update the speed, position and fitness values, and execute step (9);

[0015] (7) Sort all particles in the particle swarm according to their fitness values ​​from small to large, select the first n% of particles with small fitness values ​​and mark them as high-quality particles, mark the last m% of particles with large fitness values ​​as low-quality particles, and mark the remaining particles as ordinary particles, n, m∈[1,50]; perform mutation operation on the low-quality particles to obtain mutant particles, and the mutation formula for the mutation operation on the low-quality particles is:

[0016]

[0017]

[0018] In the formula, x i ∈[x min ,x max ], x min is the minimum map boundary value in dimension i, x max is the maximum value of the map boundary in dimension i; r is a random number in [0,1]; μ is a fixed constant; x″ i is the position of the ith particle in the mutant particle; x i is the position of the i-th particle in the particle swarm;

[0019] (8) After updating the individual optimal and global optimal values ​​of each particle in the particle swarm containing inferior particles and updating the speed, position and fitness value, the particles are combined with the mutant particles to form a new particle swarm; the fitness value of the mutant particles is calculated, and the fitness values ​​of all particles in the new particle swarm are re-sorted from small to large, and the last K particles are eliminated, where K is m% of the number of particles in the original particle swarm containing inferior particles;

[0020] (9) When the algorithm has not reached the maximum number of iterations, it returns to step (5) and performs a loop operation; when the maximum number of iterations is reached, the particle with the smallest fitness value is selected as the path solution, and the obtained solution is the optimal path solution for the robot's movement.

[0021] Beneficial effect: Compared with the prior art, the significant advantage of the present invention is that by setting a judgment number, deciding whether to perform mutation operation on inferior particles, calculating the fitness value of the mutated particles and reordering the particles, and eliminating the particles with the worst fitness value, the diversity of particles can be effectively improved, that is, the diversity of path selection in mobile robot path planning can be improved.

[0022] Furthermore, step (4) further includes, after obtaining the position of each particle, solving the symmetric position of each particle, substituting the symmetric particle position into the cost function to solve the fitness value of the symmetric particle, and saving the fitness value of the symmetric particle; comparing the fitness value of each particle with the fitness value of the symmetric particle corresponding to each particle, and replacing the original particle with the one with the smaller fitness value; the solution formula for the symmetric particle is:

[0023]

[0024] In the formula, i represents the dimension; x i ∈[a i ,b i ], that is, the size of the particle in a certain dimension, x i is the position size of the i-th particle in the particle swarm; is the symmetric position of the desired particle in a certain dimension; by solving the symmetric particles of the initial particles, it is determined which particle to select as the true initial solution, thereby accelerating the convergence speed of the particle swarm algorithm.

[0025] Furthermore, it also includes setting adaptive parameters ω, c1, c2 between step (4) and step (5);

[0026]

[0027]

[0028]

[0029]

[0030] In the formula, ω is the inertia weight, c1 is the single learning factor, and c2 is the global learning factor; d g is the distance between the particle evaluated in the current iteration and the historical global optimal particle, d min is the minimum value of the particle evaluated in history and the global optimal particle in history, d maxis the maximum value of the particle evaluated in history and the global optimal particle in history; c 1max and c 1min is the initial value and final value of c1; c 2max and c 2min is the initial value and termination value of c2. An adaptive parameter selection method is proposed to select different parameters at different stages of the algorithm, that is, according to the number of iterations, the global search is accelerated in the initial stage of the algorithm, and the local search is refined in the termination stage of the algorithm; the problem of slow convergence of the particle swarm algorithm caused by fixed parameters of the particle swarm algorithm is addressed.

[0031] Furthermore, the formulas for updating the position and velocity in step (6) and step (8) are:

[0032]

[0033]

[0034] In the formula, t represents the t-th iteration number, i represents the i-th particle; is the position of the i-th particle in the t-th iteration, is the velocity of the i-th particle in the t-th iteration; when the search space is d-dimensional, x i =(x i1 ,x i2 ,x i3 ,…,x id ), v i =(v i1 ,v i2 ,v i3 ,…,v id ); is the optimal position of the ith particle at the tth iteration, is the historical global optimal position of the entire particle swarm at the tth iteration.

[0035] Furthermore, the number of particles of the particle swarm algorithm in step (3) is 80, and the number of iterations is 50.

[0036] Furthermore, in step (8), the values ​​of n and m are both 20.

[0037] The present invention also adopts a system for mobile robot path planning, comprising:

[0038] The map construction and division module formats the robot's activity area map by using a grid map, and identifies obstacle-free areas and obstacle areas; stores the starting point and end point of the robot's movement;

[0039] The particle swarm algorithm optimization module is used to set the cost function of the particle swarm algorithm. The formula is as follows:

[0040]

[0041] Where L(p) is the shortest path obtained by the particle, L o is the distance between the mobile robot and the nearest obstacle; α and β are fixed parameters used to control the optimal path that balances path length and safety; F is the required cost function value; initialize the particle swarm algorithm parameters; set the number of particles and the total number of iterations of the particle swarm algorithm, set the particle position limit and particle speed limit conditions according to the formatted map, and randomly initialize the position and speed of each particle; after obtaining the position and speed of each particle in the particle swarm, substitute the particle position into the cost function to obtain the fitness value of each particle, and save the fitness value of each particle;

[0042] Take a random number and compare it with the decision number. The value of the random number is [0,1]. The formula of the decision number is as follows:

[0043]

[0044] In the formula, Q is the number of decisions; t is the current number of iterations, T max The maximum number of iterations set;

[0045] If the value of the random number is greater than the decision number, for each particle in the particle swarm, update the individual optimal and global optimal values ​​and update the speed, position and fitness value. When the algorithm has not reached the maximum number of iterations, return the random number and compare it with the decision number to perform a loop operation. When the maximum number of iterations is reached, select the particle with the smallest fitness value as the path solution, and the solution obtained is the optimal path solution for the robot's movement.

[0046] Otherwise, all particles in the particle swarm are sorted from small to large according to the fitness value, the first n% of particles with small fitness values ​​are selected and marked as high-quality particles, the last m% of particles with large fitness values ​​are marked as low-quality particles, and the remaining particles are marked as ordinary particles, n, m∈[1,50]; the low-quality particles are mutated to obtain mutated particles, and the mutation formula for the mutation operation of the low-quality particles is:

[0047]

[0048]

[0049] In the formula, x i ∈[x min ,x max ], x min is the minimum map boundary value in dimension i, x max is the maximum value of the map boundary in dimension i; r is a random number in [0,1]; μ is a fixed constant; x″ iis the position of the ith particle in the mutant particle; x i is the position of the ith particle in the particle swarm; after updating the individual optimal and global optimal values ​​of each particle in the particle swarm containing inferior particles and updating the speed, position and fitness value, it is combined with the mutant particles to form a new particle swarm; the fitness value of the mutant particle is calculated, and the fitness values ​​of all particles in the new particle swarm are reordered from small to large, and the K particles at the end are eliminated, where K is m% of the number of the original particle swarm containing inferior particles; when the algorithm has not reached the maximum number of iterations, the random number returned is compared with the decision number for a loop operation; when the maximum number of iterations is reached, the particle with the smallest fitness value is selected as the path solution, and the solution obtained is the optimal path solution for the robot movement.

[0050] A computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the above method.

[0051] A debugging device, a memory, a processor and a program stored and executable on the memory, wherein the program implements the steps of the above method when executed by the processor.

[0052] Beneficial effect: Compared with the prior art, the present invention has the significant advantage that the system can be used to select the optimal moving path in the path planning of the mobile robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Shown is a schematic diagram of the method for mobile robot path planning of the present invention. DETAILED DESCRIPTION

[0054] The present invention will be further described below in conjunction with the accompanying drawings.

[0055] like Figure 1 As shown, a method for mobile robot path planning according to the present invention comprises the following steps:

[0056] (1) Initialize the map of the robot's moving area;

[0057] The robot activity area map is formatted by using a grid map, where the area without obstacles in the robot activity area is represented by 0, and the area with obstacles is represented by 1; the robot is abstracted as a point, and the starting point and end point of the robot's movement are stored; the robot is generally a wheeled or tracked mobile robot by default.

[0058] (2) Considering the length and smoothness of the robot's moving path, the cost function of the particle swarm algorithm is set as follows:

[0059]

[0060] Where L(p) is the shortest path obtained by the particle, L o is the distance between the mobile robot and the nearest obstacle; α and β are fixed parameters used to control the optimal path that balances path length and safety; F is the required cost function value;

[0061] (3) Set the number of particles and the total number of iterations of the particle swarm algorithm, the number of particles is 80, and the number of iterations is 50; initialize the parameters of the particle swarm algorithm; set the position restriction conditions of the particles in the particle swarm based on the formatted map obtained in step (1), which cannot exceed the boundary of the map; set the speed restriction conditions of the particles in the particle swarm, and the speed is set to one tenth of the particle position; randomly initialize the position and speed of each particle;

[0062] (4) After obtaining the position of each particle in the particle swarm, substitute the position of the particle into the cost function to obtain the fitness value of each particle, and save the fitness value of each particle; according to the position of each particle in the particle swarm, solve the symmetric position of each particle, substitute the position of the symmetric particle into the cost function to obtain the fitness value of the symmetric particle, and save the fitness value of the symmetric particle; compare the fitness value of each particle with the fitness value of the symmetric particle corresponding to each particle, and take the one with the smaller fitness value to replace the original particle;

[0063] The solution formula for the symmetric particle is:

[0064]

[0065] In the formula, i represents the dimension; x i ∈[a i ,b i ], that is, the size of the particle in a certain dimension, x i is the position size of the particle in the particle group; is the symmetric position of the required particle in a certain dimension.

[0066] (5) Setting adaptive parameters ω, c1, c2;

[0067]

[0068]

[0069]

[0070]

[0071] In the formula, ω is the inertia weight, c1 is the single learning factor, and c2 is the global learning factor; d g is the distance between the particle evaluated in the current iteration and the historical global optimal particle, d minis the minimum value of the particle evaluated in history and the global optimal particle in history, d max is the maximum value of the particle evaluated in history and the global optimal particle in history; c 1max and c 1min is the initial value and final value of c1; c 2max and c 2min is the initial value and final value of c2;

[0072] (6) Take a random number and compare it with the decision number. If the value of the random number is greater than the decision number, execute step (6); otherwise, execute step (7). The value of the random number is [0, 1]. The formula of the decision number is as follows:

[0073]

[0074] In the formula, Q is the number of decisions; t is the current number of iterations, T max The maximum number of iterations set;

[0075] (7) For each particle in the particle swarm, update the individual optimal value, the global optimal value, and update the speed, position, and fitness value, and execute step (9); the individual optimal value refers to the optimal speed and position of a single particle in the particle swarm, and the global optimal value refers to comparing the speed and position of each particle in the particle swarm, and obtaining the optimal position and its corresponding speed as the global optimal value; the formula for updating the speed and position is:

[0076]

[0077]

[0078] In the formula, t represents the t-th iteration number, i represents the i-th particle; is the position of the i-th particle in the t-th iteration, is the velocity of the i-th particle in the t-th iteration; when the search space is d-dimensional, x i =(x i1 ,x i2 ,x i3 ,…,x id ), v i =(v i1 ,v i2 ,v i3 ,…,v id ); is the optimal position of the ith particle at the tth iteration, is the historical global optimal position of the entire particle swarm at the tth iteration.

[0079] (8) Sort all particles in the particle swarm according to their fitness values ​​from small to large, select the first 20% of particles with small fitness values ​​and mark them as high-quality particles, the last 20% of particles with large fitness values ​​and mark them as low-quality particles, and the remaining particles are marked as ordinary particles, n, m ∈ [1,50];

[0080] A mutation operation is performed on the inferior particles in the particle group to obtain mutant particles. The mutation formula for the mutation operation on the inferior particles is:

[0081]

[0082]

[0083] In the formula, x i ∈[x min ,x max ], x min is the minimum map boundary value in dimension i, x max is the maximum value of the map boundary in dimension i; r is a random number in [0,1]; μ is a fixed constant; x″ i is the position of the ith particle in the mutant particle; x i is the position of the i-th particle in the particle swarm;

[0084] (9) After updating the individual optimal and global optimal values ​​of each particle in the original particle group containing inferior particles and updating the speed, position and fitness value, the individual optimal refers to the optimal speed and position of a single particle in the particle group, and the global optimal refers to comparing the speed and position of each particle in the particle group, and obtaining the optimal position and its corresponding speed as the global optimal value; the formula for updating the position and speed is the above calculation formula (8);

[0085] The original particle group and the mutant particles are combined to form a new particle group; the fitness value of the mutant particles is calculated, and the fitness values ​​of all particles in the new particle group are reordered from small to large, and the last K particles are eliminated, where K is 20% of the number of the original particle group containing inferior particles;

[0086] (10) When the algorithm has not reached the maximum number of iterations, it returns to step (6) and performs a loop operation; when the maximum number of iterations is reached, the particle with the smallest fitness value is selected as the path solution, and the obtained solution is the optimal path solution for the robot's movement.

[0087] In addition, the present invention also provides a system for mobile robot path planning, comprising:

[0088] The map construction and division module formats the robot's activity area map by using a grid map. The area without obstacles in the robot's activity area is represented by 0, and the area with obstacles is represented by 1; the starting point and end point of the robot's movement are stored;

[0089] The particle swarm algorithm optimization module is used to consider the robot's moving path length and path smoothness, and set the particle swarm algorithm cost function. The formula is as follows:

[0090]

[0091] Where L(p) is the shortest path obtained by the particle, L o is the distance between the mobile robot and the nearest obstacle; α and β are fixed parameters used to control the optimal path that balances path length and safety; F is the required cost function value; initialize the particle swarm algorithm parameters; set the number of particles and the total number of iterations of the particle swarm algorithm, set the particle position limit and particle speed limit conditions according to the formatted map, and randomly initialize the position and speed of each particle; after obtaining the position and speed of each particle in the particle swarm, substitute the particle position into the cost function to obtain the fitness value of each particle, and save the fitness value of each particle;

[0092] Take a random number and compare it with the decision number. The value of the random number is [0,1]. The formula of the decision number is as follows:

[0093]

[0094] In the formula, Q is the number of decisions; t is the current number of iterations, T max The maximum number of iterations set;

[0095] If the value of the random number is greater than the decision number, for each particle in the particle swarm, update the individual optimal and global optimal values ​​and update the speed, position and fitness value. When the algorithm has not reached the maximum number of iterations, return the random number and compare it with the decision number to perform a loop operation. When the maximum number of iterations is reached, select the particle with the smallest fitness value as the path solution, and the solution obtained is the optimal path solution for the robot's movement.

[0096] Otherwise, all particles in the particle swarm are sorted from small to large according to the fitness value, the first n% of particles with small fitness values ​​are selected and marked as high-quality particles, the last m% of particles with large fitness values ​​are marked as low-quality particles, and the remaining particles are marked as ordinary particles, n, m∈[1,50]; the low-quality particles are mutated to obtain mutated particles, and the mutation formula for the mutation operation of the low-quality particles is:

[0097]

[0098]

[0099] In the formula, x i ∈[x min ,x max ], x minis the minimum map boundary value in dimension i, x max is the maximum value of the map boundary in dimension i; r is a random number in [0,1]; μ is a fixed constant; x″ i is the position of the ith particle in the mutant particle; x i is the position of the ith particle in the particle swarm; after updating the individual optimal and global optimal values ​​of each particle in the particle swarm containing inferior particles and updating the speed, position and fitness value, it is combined with the mutant particles to form a new particle swarm; the fitness value of the mutant particle is calculated, and the fitness values ​​of all particles in the new particle swarm are reordered from small to large, and the K particles at the end are eliminated, where K is m% of the number of the original particle swarm containing inferior particles; when the algorithm has not reached the maximum number of iterations, the random number returned is compared with the decision number for a loop operation; when the maximum number of iterations is reached, the particle with the smallest fitness value is selected as the path solution, and the solution obtained is the optimal path solution for the robot movement.

[0100] The present invention also provides a computer-readable storage medium, characterized in that the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned method of mobile robot path planning.

[0101] The present invention also provides a debugging device, characterized by a memory, a processor and a program stored and executable on the memory, wherein the program implements the steps of the above-mentioned method for mobile robot path planning when executed by the processor.

Claims

1. A method for mobile robot path planning, characterized in that: The following steps are involved: (1) Formatting the robot's activity area map by using a grid map and identifying obstacle-free areas and obstacle-affected areas; storing the starting point and end point of the robot's movement; (2) Set the cost function of the particle swarm algorithm, the formula is as follows: Where L(p) is the shortest path obtained by the particle, L o is the distance between the mobile robot and the nearest obstacle; α and β are fixed parameters used to control the optimal path that balances path length and safety; F is the required cost function value; (3) Initialize the parameters of the particle swarm algorithm; set the number of particles and the total number of iterations of the particle swarm algorithm, set the particle position limit and particle speed limit conditions according to the formatted map, and randomly initialize the position and speed of each particle; (4) After obtaining the position and velocity of each particle in the particle swarm, substitute the particle position into the cost function to obtain the fitness value of each particle, and save the fitness value of each particle; (5) Take a random number and compare it with the decision number. If the value of the random number is greater than the decision number, execute step (6); otherwise, execute step (7). The value of the random number is [0, 1]. The formula of the decision number is as follows: In the formula, Q is the number of decisions; t is the current number of iterations, T max The maximum number of iterations set; (6) For each particle in the particle swarm, update the individual optimal and global optimal values ​​and update the speed, position and fitness values, and execute step (9); (7) Sort all particles in the particle swarm according to their fitness values ​​from small to large, select the first n% of particles with small fitness values ​​and mark them as high-quality particles, mark the last m% of particles with large fitness values ​​as low-quality particles, and mark the remaining particles as ordinary particles, n, m∈[1,50]; perform mutation operation on the low-quality particles to obtain mutant particles, and the mutation formula for performing mutation operation on the low-quality particles is: In the formula, x i ∈[x min ,x max ], x min is the minimum map boundary value in dimension i, x max is the maximum value of the map boundary in dimension i; r is a random number in [0,1]; μ is a fixed constant; x″ i is the position of the ith particle in the mutant particle; x i is the position of the i-th particle in the particle swarm; (8) After updating the individual optimal and global optimal values ​​of each particle in the particle swarm containing inferior particles and updating the speed, position and fitness value, the particles are combined with the mutant particles to form a new particle swarm; the fitness value of the mutant particles is calculated, and the fitness values ​​of all particles in the new particle swarm are re-sorted from small to large, and the last K particles are eliminated, where K is m% of the number of particles in the original particle swarm containing inferior particles; (9) When the algorithm has not reached the maximum number of iterations, return to step (5) and perform a loop operation; when the maximum number of iterations is reached, the particle with the smallest fitness value is selected as the path solution, and the obtained solution is the optimal path solution for the robot to move; Step (4) also includes, after obtaining the position of each particle, solving the symmetric position of each particle, substituting the position of the symmetric particle into the cost function to solve the fitness value of the symmetric particle, and saving the fitness value of the symmetric particle; comparing the fitness value of each particle with the fitness value of the symmetric particle corresponding to each particle, and replacing the original particle with the one with the smaller fitness value; the solution formula of the symmetric particle is: In the formula, i represents the dimension; x i ∈[a i ,b i ], that is, the size of the particle in a certain dimension, x i is the position and size of the particles in the particle group; is the symmetric position of the desired particle in a certain dimension; It also includes setting adaptive parameters ω, c1, c2 between step (4) and step (5); In the formula, ω is the inertia weight, c1 is the single learning factor, and c2 is the global learning factor; d g is the distance between the particle evaluated in the current iteration and the historical global optimal particle, d min is the minimum value of the particle evaluated in history and the global optimal particle in history, d max is the maximum value of the particle evaluated in history and the global optimal particle in history; c 1max and c 1min is the initial value and final value of c1; c 2max and c 2min are the initial and final values ​​of c2.

2. The method for mobile robot path planning according to claim 1, characterized in that: The formulas for updating position and velocity in steps (6) and (8) are: In the formula, t represents the t-th iteration number, i represents the i-th particle; is the position of the i-th particle in the t-th iteration, is the velocity of the i-th particle in the t-th iteration; when the search space is d-dimensional, x i =(x i1 ,x i2 ,x i3 ,…,x id ), v i =(v i1 ,v i2 ,v i3 ,…,v id ); is the optimal position of the ith particle at the tth iteration, is the historical global optimal position of the entire particle swarm at the tth iteration.

3. The method for mobile robot path planning according to claim 1, characterized in that: The number of particles of the particle swarm algorithm in step (3) is 80, and the total number of iterations is 50.

4. The method for mobile robot path planning according to claim 1, characterized in that: In step (7), the values ​​of n and m are both 20.

5. A system for mobile robot path planning using the method of claim 1, characterized in that: include: The map construction and division module formats the robot's activity area map by using a grid map and identifies obstacle-free areas and obstacle-filled areas; stores the starting and ending points of the robot's movement; The particle swarm algorithm optimization module is used to set the cost function of the particle swarm algorithm. The formula is as follows: Where L(p) is the shortest path obtained by the particle, L o is the distance between the mobile robot and the nearest obstacle; α and β are fixed parameters used to control the optimal path that balances path length and safety; F is the required cost function value; initialize the particle swarm algorithm parameters; set the number of particles and the total number of iterations of the particle swarm algorithm, set the particle position limit and particle speed limit conditions according to the formatted map, and randomly initialize the position and speed of each particle; after obtaining the position and speed of each particle in the particle swarm, substitute the particle position into the cost function to obtain the fitness value of each particle, and save the fitness value of each particle; The method also includes solving the symmetric position of each particle after obtaining the position of each particle, substituting the position of the symmetric particle into the cost function to solve the fitness value of the symmetric particle, and saving the fitness value of the symmetric particle; comparing the fitness value of each particle with the fitness value of the symmetric particle corresponding to each particle, and replacing the original particle with the one with the smaller fitness value; the solution formula of the symmetric particle is: In the formula, i represents the dimension; x i ∈[a i ,b i ], that is, the size of the particle in a certain dimension, x i is the position and size of the particles in the particle group; is the symmetric position of the desired particle in a certain dimension; Set adaptive parameters ω, c1, c2; In the formula, ω is the inertia weight, c1 is the single learning factor, and c2 is the global learning factor; d g is the distance between the particle evaluated in the current iteration and the historical global optimal particle, d min is the minimum value of the particle evaluated in history and the global optimal particle in history, d max is the maximum value of the particle evaluated in history and the global optimal particle in history; c 1max and c 1min is the initial value and final value of c1; c 2max and c 2min is the initial value and final value of c2; Take a random number and compare it with the decision number. The value of the random number is [0,1]. The formula of the decision number is as follows: In the formula, Q is the number of decisions; t is the current number of iterations, T max The maximum number of iterations set; If the value of the random number is greater than the decision number, for each particle in the particle swarm, update the individual optimal and global optimal values ​​and update the speed, position and fitness value. When the algorithm has not reached the maximum number of iterations, return the random number and compare it with the decision number to perform a loop operation. When the maximum number of iterations is reached, select the particle with the smallest fitness value as the path solution, and the solution obtained is the optimal path solution for the robot's movement. Otherwise, all particles in the particle swarm are sorted from small to large according to the fitness value, the first n% of particles with small fitness values ​​are selected and marked as high-quality particles, the last m% of particles with large fitness values ​​are marked as low-quality particles, and the remaining particles are marked as ordinary particles, n, m∈[1,50]; the low-quality particles are mutated to obtain mutant particles, and the mutation formula for the mutation operation on the low-quality particles is: In the formula, x i ∈[x min ,x max ], x min is the minimum map boundary value in dimension i, x max is the maximum value of the map boundary in dimension i; r is a random number in [0,1]; μ is a fixed constant; x″ i is the position of the ith particle in the mutant particle; x i is the position of the ith particle in the particle swarm; after updating the individual optimal and global optimal values ​​of each particle in the particle swarm containing inferior particles and updating the speed, position and fitness value, it is combined with the mutant particles to form a new particle swarm; the fitness value of the mutant particle is calculated, and the fitness values ​​of all particles in the new particle swarm are reordered from small to large, and the K particles at the end are eliminated, where K is m% of the number of the original particle swarm containing inferior particles; when the algorithm has not reached the maximum number of iterations, the random number returned is compared with the decision number for a loop operation; when the maximum number of iterations is reached, the particle with the smallest fitness value is selected as the path solution, and the solution obtained is the optimal path solution for the robot movement.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 4.

7. A debugging device, characterized in that: A memory, a processor, and a program stored and executable on the memory, wherein when the program is executed by the processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images