Robot optimal path planning method based on improved angle exendin optimization algorithm

Abstract

The invention discloses a robot optimal path planning method based on an improved angle skid optimization algorithm, and the method comprises the steps: designing a single-target optimization problem under a constraint condition based on the angle skid optimization algorithm, setting a target function as the shortest distance between a starting point and an end point, and determining the constraint condition; a good point set strategy is adopted to replace random node initialization distribution, and an adaptive dynamic weight, a dynamic weight, an adaptive coefficient, a Brownian motion random disturbance parameter and a random walk and adaptive t distribution disturbance strategy of step length adjustment are introduced. A hiding behavior, a blood jetting behavior, an escape behavior, a skin blackening and brightening behavior and a melanocyte stimulating hormone (MSH) rate formula of the algorithm are optimized respectively, and a planned path is optimized through the improved HLOA algorithm. Compared with the prior art, the method has the advantages that the convergence speed, the optimization precision and the global search capability of the angle exid optimization algorithm in robot path planning are improved, and the method is suitable for optimal path planning in a complex environment.

Description

Technical Field

[0001] The present invention relates to the field of path planning design, a robot optimal path planning method based on an improved horned lizard optimization algorithm. Background Art

[0002] In the field of modern industrial and service robots, path planning is a key technology that determines whether the robot can perform tasks efficiently and safely in complex environments. The core of the path planning problem is to find the optimal path from the starting point to the end point, so that the robot can avoid obstacles and meet various constraints, such as the shortest path length, the least energy consumption or the shortest time.

[0003] Traditional path planning methods include A* algorithm, Dijkstra algorithm and rapidly expanding random tree (RRT). However, these methods have certain limitations when dealing with complex high-dimensional space and dynamic environment, such as large amount of calculation, easy to fall into local optimum, poor adaptability to environmental changes, etc. Therefore, studying more efficient and robust path planning algorithms has become one of the important directions in the field of robotics research.

[0004] Swarm intelligence algorithms have been widely used in optimization problems in recent years due to their powerful global search capabilities and adaptive characteristics. Among them, classic swarm intelligence algorithms such as particle swarm optimization (PSO) and ant colony optimization (ACO) have been successfully applied to path planning problems. However, these algorithms still have some shortcomings, such as slow convergence speed, easy to fall into local optimal solutions, and complex parameter adjustment.

[0005] Horned Lizard Optimization Algorithm (HLOA) is an emerging swarm intelligence algorithm, which is inspired by the hiding, darkening or lightening of skin, blood spraying and movement-escape defense behaviors of horned lizards in nature. HLOA algorithm has the advantages of simple structure, easy implementation and strong search ability.

[0006] Although HLOA has demonstrated its superiority in solving some optimization problems, its application in robot path planning is still in its early stages. Existing HLOA algorithms may face problems such as slow convergence and inaccurate path planning when planning paths in complex environments. Therefore, how to improve the HLOA algorithm to improve its performance in robot path planning has become an important and challenging research topic. Summary of the invention

[0007] Purpose of the invention: In view of the shortcomings of the prior art, the present invention proposes a robot optimal path planning method based on an improved horned lizard optimization algorithm. By introducing a new disturbance mechanism and an adaptive adjustment strategy, the improved HLOA algorithm achieves a better balance between global search and local search capabilities, solves the problems of slow convergence speed, easy to fall into local optimal solutions, and inaccurate path planning in existing path planning algorithms in complex environments, and provides more reliable technical support for robot autonomous navigation.

[0008] Technical solution: The present invention discloses a robot optimal path planning method based on an improved horned lizard optimization algorithm, comprising the following steps:

[0009] Step 1: Design a single-objective optimization problem based on the horned lizard optimization algorithm to solve the constraint conditions. In the planned candidate path, the objective function is set as the shortest distance between the starting point and the end point, and the following constraints must be met: the path must not intersect with obstacles, and the minimum distance between each node and the obstacle must be greater than zero; the constraints are integrated into the objective function as part of the horned lizard population fitness function to update the fitness value of the population;

[0010] Step 2: Use the optimal point set method to distribute the horned lizard population in the search space, where each horned lizard represents an independent candidate solution, and each candidate solution represents a possible path node. The fitness value is updated through continuous iteration, and the better solution is determined by the size of the ranking to optimize the planned path;

[0011] Step 3: According to the fitness value ranking information obtained from the search, the path with the lowest fitness value is selected as the candidate path to be planned; the behavior of the horned lizard population at different stages is improved, and some horned lizard populations are updated through adaptive dynamic weights and random perturbation strategies;

[0012] Step 4: When the objective function is satisfied, the position of the horned lizard individual will be converted into a path node; the optimal solution of the problem is converted into numerical coordinates, each coordinate corresponds to the position of a path node, and the numerical coordinates are mapped to the actual coordinate space to generate the path node; the path node is normalized and the path between the starting point and the end point is generated by linear interpolation.

[0013] Furthermore, in step 2, the optimal point set method is used to distribute the horned lizard population, as follows:

[0014] Assume Hs is the unit cube of s-dimensional Euclidean space, in which there exists a set of good points P n (k), as shown in the following formula:

[0015]

[0016] Where n represents the number of samples, k represents the kth horned lizard individual, and r represents an optimal point, which is expressed by the following formula:

[0017]

[0018] Among them, q is satisfied The smallest prime number of , i means the variable dimension is the i-th dimension;

[0019] Mapping the good point set to the feasible region where the population is located, we get:

[0020]

[0021] in, represents the position of the kth horned lizard individual in dimension i, ub represents the upper bound of the current dimension, and lb represents the lower bound of the current dimension.

[0022] Furthermore, in step 3, the behavior of horned lizard populations at different stages is improved, specifically including:

[0023] Add an adaptive dynamic weight q to the hidden behavior of the horned lizard optimization algorithm. The formula is as follows:

[0024]

[0025] Among them, maxIdx represents the current worst solution, minIdx represents the current optimal solution, t represents the current number of iterations, Max_iter represents the maximum number of iterations, α and β represent two weight values. As the algorithm iterates, The value gradually increases, The value gradually decreases;

[0026] Substituting the adaptive dynamic weight q into the hidden behavior formula, we get:

[0027]

[0028] Among them, x i (t+1) is the position of the horned lizard in the solution search space of the t+1th generation, x best (t) is the optimal solution of generation t; r1, r2, r3 and r4 are integer random numbers generated between 1 and the maximum size of the population, where r1≠r2≠r3≠r4; and is the selected search solution r1, r2, r3, r4; σ is a binary value, Set to 2, c1, c2 are random numbers drawn from Table 29 containing a standardized palette, and c1 ≠ c2.

[0029] Furthermore, the improvement of the behavior of horned lizard populations at different stages in step 3 also includes:

[0030] The blood ejection behavior formula of the horned lizard optimization algorithm is modified by replacing Replaced with the dynamic weight ω, the formula is as follows: t represents the current iteration number, Max_iter represents the maximum iteration number; the dynamic weight ω is introduced into the blood ejection behavior formula to obtain:

[0031]

[0032] Among them, x i (t+1) is the position of the horned lizard in the solution search space of the t+1th generation, x best (t) is the optimal solution of generation t, x i (t) is the position of the horned lizard in the solution search space of the tth generation, v0 is set to 1, and δ is set to g is the earth's gravity, ε is set to 10 -6 .

[0033] Furthermore, the improvement of the behavior of horned lizard populations at different stages in step 3 also includes:

[0034] Modify the escape behavior formula of the horned lizard optimization algorithm and add an adaptive coefficient As shown below:

[0035]

[0036] Among them, x i (t+1) is the position of the horned lizard in the solution search space of the t+1th generation, x best (t) is the optimal solution of generation t, x i (t) is the position of the horned lizard in the solution search space of the tth generation, Expressed as walk represents the generation of random numbers between -1 and 1, and γ is the mean of random numbers generated by the standard Cauchy distribution.

[0037] Furthermore, the improvement of the behavior of horned lizard populations at different stages in step 3 also includes:

[0038] The skin darkening and lightening behavior formulas of the horned lizard optimization algorithm are modified by adding the Brownian motion random perturbation parameter η; the skin lightening behavior is shown in the following formula:

[0039]

[0040] The skin lightening behavior of the horned lizard is shown in the following formula:

[0041]

[0042] Where σ is a random binary value. If the value of σ is 1, the skin lightening behavior formula of the horned lizard is executed. If the value of σ is 0, the skin darkening behavior formula of the horned lizard is executed. best (t) is the optimal solution of generation t, Light1 and Light2 are random numbers generated between 0 and 0.4046661, Dark1 and Dark2 are random numbers generated between 0.5440510 and 1, and x worst (t) represents the worst search result value, r1, r2, r3 and r4 are integer random numbers generated between 1 and the maximum number of the population, where r1≠r2≠r3≠r4; and are the selected search solutions r1, r2, r3, and r4.

[0043] Furthermore, the improvement of the behavior of horned lizard populations at different stages in step 3 also includes:

[0044] The melanocyte stimulating hormone (MSH) rate formula of the horned lizard optimization algorithm was modified by using a random walk with step size adjustment and an adaptive t-distribution perturbation strategy. The formula is as follows:

[0045]

[0046] x r (t) = x best (t)+C_iter(t)·x best (t)

[0047] Among them, x r (t) is the position of the optimal solution after the adaptive t-distribution mutation disturbance after the t-th iteration of the algorithm; x i (t) is the position of the horned lizard in the solution search space of the tth generation, x best (t) is the optimal solution of generation t, C_iter(t) is the degree of freedom parameter of the distribution with the number of iterations as t, and stepsize represents the step size parameter of the random walk.

[0048] Beneficial effects:

[0049] The present invention adopts the optimal point set strategy to replace the traditional random node initialization distribution, so that the distribution of horned lizard population in the search space is more uniform, the global exploration capability is improved, and the risk of falling into the local optimum is reduced.

[0050] The present invention uses three different adaptive dynamic weight strategies to improve and replace the behavior formulas for hiding behavior, blood spraying behavior, and mobile escape behavior. Specifically, the adaptive dynamic weight q is introduced to enable the algorithm to automatically adjust the weight at different stages to optimize the optimization efficiency. At the same time, the dynamic weight ω is used to replace the original sin function, making the algorithm more flexible in direction adjustment, thereby improving the accuracy and stability of path planning. In addition, the added adaptive coefficient This enables the algorithm to better balance exploration and development, enhancing overall performance and robustness.

[0051] The Brownian motion random perturbation parameter η introduced into the skin darkening or lightening behavior in the horned lizard optimization algorithm of the present invention further increases the randomness and diversity of the algorithm, avoids falling into local optimality, and improves the global search capability.

[0052] The present invention adopts an improvement measure combining a random walk with adaptive step size adjustment and an adaptive t-distribution strategy for the melanocyte stimulating hormone (MSH) rate behavior in the horned lizard optimization algorithm. The random walk with step size adjustment and the adaptive t-distribution perturbation strategy enable the algorithm to more accurately approach the global optimal solution during the convergence process, thereby significantly improving the path planning effect. In summary, the present invention significantly improves the convergence speed, optimization accuracy and global search capability of the horned lizard optimization algorithm when performing robot optimal path planning in complex environments through a number of improvements. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Optimize the algorithm flow chart for the improved horned lizard;

[0054] Figure 2 Initialize the population for the good point set and the scatter plot of the randomly initialized population;

[0055] Figure 3 The tracking path diagram of the improved horned lizard optimization algorithm and the original algorithm in the environment of 40*40 grid map;

[0056] Figure 4 Partial iteration graphs for solving 23 classic functions using the improved horned lizard optimization algorithm and other intelligent algorithms. DETAILED DESCRIPTION

[0057] The present invention will be described in detail below with reference to the embodiments shown in the accompanying drawings.

[0058] refer to Figure 1 , a robot optimal path planning method based on an improved horned lizard optimization algorithm disclosed in the present invention is described in detail, comprising the following steps:

[0059] Step 1: In the search space, the best point set strategy is used to replace the random node initialization distribution of the horned lizard population, where each horned lizard represents an independent candidate solution, and each candidate solution represents a possible path node. These horned lizards update their fitness values ​​through continuous iterations, and determine the better solution by the size of the ranking. The design is based on the horned lizard optimization algorithm to solve the single-objective optimization problem under constraints, so as to optimize the planned path.

[0060] The good point set is a theory proposed by Chinese mathematician Hua Luogeng. Assume that Hs is a unit cube in s-dimensional Euclidean space, in which there exists a good point set P n (k), as shown in the following formula:

[0061]

[0062] Where n represents the number of samples, k represents the kth horned lizard individual, and r represents an optimal point, which is expressed by the following formula:

[0063]

[0064] Among them, q is satisfied The smallest prime number, i means the variable dimension is in the i-th dimension.

[0065] Mapping the good point set to the feasible region where the population is located, we get:

[0066]

[0067] in, represents the position of the kth horned lizard individual in dimension i, ub represents the upper bound of the current dimension, and lb represents the lower bound of the current dimension.

[0068] Initialize the original random population function formula Replaced with the population formula generated by initialization using the good point set strategy pass Figure 2 It can be found that the randomly generated population is unevenly distributed in the entire solution space. It is very concentrated in some areas and very scattered in other areas, resulting in low utilization of the entire search space and weak population diversity. The initialization of the good point set is more evenly distributed than random nodes. This makes the algorithm more traversable and helps the algorithm achieve the best global optimization effect.

[0069] Step 2: According to the fitness value ranking information obtained from the search, the path with the lowest fitness value is selected as the candidate path to be planned. According to the behavior of horned lizard populations at different stages, some horned lizard populations are updated through strategies such as adaptive dynamic weights and random disturbances.

[0070] (1) Add an adaptive dynamic weight q to the hidden behavior of the horned lizard optimization algorithm. The formula is as follows:

[0071]

[0072] Where maxIdx represents the current worst solution, minIdx represents the current optimal solution, t represents the current number of iterations, Max_iter represents the maximum number of iterations, and α and β represent two weight values. As the algorithm iterates, The value gradually increases, The value decreases gradually.

[0073] In the early stage of algorithm iteration, if the horned lizard population falls into a local optimal solution, the difference between the optimal solution and the worst solution is very small, that is, the value of maxIdx-minIdx is very small, but The design of this part can still obtain a larger weight value q, which prevents the algorithm from falling into a small-range search at the beginning of the iteration. As the number of iterations of the horned lizard population increases, if the algorithm has not obtained the optimal solution, the value of maxIdx-minIdx plays a dominant role in the weight value q, allowing the algorithm to find the optimal solution with a larger step size. This design of adaptively adjusting the first half of the weight q plays a major regulatory role when the number of population iterations is too large, and the second half plays a major regulatory role when the population falls into a local optimum. The front and back parts of the weight q will change according to the changes in the current population position, and are highly adaptive. Substituting it into the hidden behavior formula, we get:

[0074]

[0075] Among them, x i (t+1) is the position of the horned lizard in the solution search space of the t+1th generation, x best (t) is the optimal solution of generation t; r1, r2, r3 and r4 are integer random numbers generated between 1 and the maximum size of the population, where r1≠r2≠r3≠r4; and is the selected search solution of r1, r2, r3, r4. Max_iter represents the maximum number of iterations, σ is a binary value, Set to 2, c1, c2 are random numbers drawn from Table 29 containing a standardized palette, and c1 ≠ c2.

[0076] (2) The blood ejection behavior formula of the horned lizard optimization algorithm was modified by replacing Replaced with the dynamic weight ω, the formula is as follows:

[0077]

[0078] Substituting ω into the blood ejection behavior formula, we get:

[0079]

[0080] Among them, v0 is set to 1, δ is set to g is the Earth's gravity, set to 0.00987km / s 2 ,ε is set to 10 -6 The function that plays a decisive role in the output result in the original blood ejection behavior formula is and The mathematical properties of these two functions are that they decrease from 1 to 0 as the number of iterations increases, and they affect the optimal price x of the current algebra respectively. best (t) and the current solution x i (t). However, we hope to pay more attention to the results of local search in the later stages of algorithm iteration, that is, to adjust the function in the formula ω is a nonlinear increasing state from 0 to 1 with the iteration of the algorithm, which is in line with the expected effect.

[0081] (3) Modify the escape behavior formula of the horned lizard optimization algorithm and add an adaptive coefficient As shown below:

[0082]

[0083] in, Expressed as Walk represents a random number generated between -1 and 1, and γ is the average value of random numbers generated by the standard Cauchy distribution. The value of shows a decreasing trend from 1 to 0 with the iteration of the algorithm, gradually reducing the weight of the optimal solution to prevent the algorithm from falling into a local optimal solution in the later stage, making it impossible to develop a more effective solution.

[0084] (4) The skin darkening and lightening behavior formulas of the horned lizard optimization algorithm are modified by adding the Brownian motion random perturbation parameter η.

[0085] The skin lightening behavior is expressed in the following formula:

[0086]

[0087] The skin lightening behavior of the horned lizard is shown in the following formula:

[0088]

[0089] Where σ is a random binary value. If the value of σ is 1, the skin lightening behavior formula of the horned lizard is executed. If the value of σ is 0, the skin darkening behavior formula of the horned lizard is executed. Light1 and Light2 are random numbers generated between 0 and 0.4046661, and Dark1 and Dark2 are random numbers generated between 0.5440510 and 1.worst (t) represents the worst search result value. The randomness of Brownian motion can help the algorithm avoid falling into the local optimal solution during the search process, increasing the algorithm's exploration ability. By increasing the random perturbation parameter η, the algorithm can search the solution space more widely, thereby increasing the probability of finding the global optimal solution.

[0090] (5) The melanocyte stimulating hormone (MSH) rate formula of the horned lizard optimization algorithm was modified by adopting a random walk with step size adjustment and an adaptive t-distribution perturbation strategy. The formula is as follows:

[0091]

[0092] x r (t) = x best (t)+C_iter(t)·x best (t)

[0093] Among them, the original formula In Replace with x r (t), x r (t) is the position of the optimal solution after the tth iteration of the algorithm after the adaptive t-distribution mutation perturbation; C_iter(t) is the degree of freedom parameter of the t-distribution with the number of iterations. The t-distribution mutation operator is used to perturb the position of the solution, so that the algorithm has better global development capabilities in the early stage of iteration and good local exploration capabilities in the later stage of iteration, and improves the convergence speed of the algorithm.

[0094] stepsize represents the random walk step size parameter. If the number of iterations is less than the set mid-iter, the step size is expressed as stepsize = initial_stepsize; otherwise, the step size is expressed as If the number of iterations is greater than the maximum number of iterations minus the number of mid-iterations, the step size

[0095] Wherein initial_stepsize represents the initial step size, decay_rate1 represents the decay rate of rapid decline in the middle stage, decay_rate2 represents the decay rate of the low step size in the final stage, and mid_iter=Max_iter / 2 represents the iteration stage of rapid decline in the middle stage.

[0096] Keeping the step size of the random walk high in the early stages of the algorithm iteration can increase the ability to explore the search space, which helps the algorithm escape from the local optimal solution and search the solution space more widely. Rapidly reducing the step size in the mid-term allows the algorithm to gradually reduce the search range when it is close to the optimal solution. This step size adjustment allows the algorithm to search the solution space more finely, thereby converging to the optimal solution more quickly. In the later stages of the iteration, keeping the step size low helps prevent the algorithm from oscillating near the optimal solution and avoids converging to a suboptimal solution too early. A low step size allows the algorithm to make fine adjustments near the optimal solution to ensure that the global optimal solution is found.

[0097] Step 3: In the planned candidate path, the objective function is set as the shortest distance between the starting point and the end point, and the following constraints must be met: the path must not intersect with obstacles, and the minimum distance between each node and the obstacle must be greater than zero. These constraints are integrated into the objective function as part of the fitness function of the horned lizard population and are used to update the fitness value of the population.

[0098] Step 4: When the objective function is satisfied, the position of the horned lizard individual will be converted into a path node.

[0099] First, the optimal solution to the problem is converted into numerical coordinates, each of which corresponds to the location of a path node. Then, these numerical coordinates are mapped to the actual coordinate space to generate path nodes. Finally, these nodes are normalized and linearly interpolated to generate the path between the starting point and the end point.

[0100] Figure 3 It can be observed that the improved horned lizard optimization algorithm IHLOA of the present invention has significantly improved the convergence speed, optimization accuracy and global search ability of the horned lizard optimization algorithm HLOA when performing optimal path planning for robots in complex environments through multiple improvements. At the same time, the improved algorithm is compared with the original algorithm and other algorithms in classic test functions. Figure 4 It can be found that the algorithm has good adaptability and robustness in many different types of optimization problems.

[0101] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent transformation or modification made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. A robot optimal path planning method based on an improved horned lizard optimization algorithm, characterized in that: The following steps are involved: Step 1: Design a single-objective optimization problem based on the horned lizard optimization algorithm to solve the constraint conditions. In the planned candidate path, the objective function is set as the shortest distance between the starting point and the end point, and the following constraints must be met: the path must not intersect with obstacles, and the minimum distance between each node and the obstacle must be greater than zero; the constraints are integrated into the objective function as part of the horned lizard population fitness function to update the fitness value of the population; Step 2: Use the optimal point set method to distribute the horned lizard population in the search space, where each horned lizard represents an independent candidate solution, and each candidate solution represents a possible path node. The fitness value is updated through continuous iteration, and the better solution is determined by the size of the ranking to optimize the planned path; Step 3: According to the fitness value ranking information obtained from the search, the path with the lowest fitness value is selected as the candidate path to be planned; the behavior of the horned lizard population at different stages is improved, and some horned lizard populations are updated through adaptive dynamic weights and random perturbation strategies; Step 4: When the objective function is satisfied, the position of the horned lizard individual will be converted into a path node; the optimal solution of the problem is converted into numerical coordinates, each coordinate corresponds to the position of a path node, and the numerical coordinates are mapped to the actual coordinate space to generate the path node; The path nodes are normalized and the path between the start point and the end point is generated by linear interpolation.

2. The robot optimal path planning method based on the improved horned lizard optimization algorithm according to claim 1 is characterized in that: In step 2, the optimal point set method is used to distribute the horned lizard population, as follows: Assume Hs is the unit cube of s-dimensional Euclidean space, in which there exists a set of good points P n (k), as shown in the following formula: Where n represents the number of samples, k represents the kth horned lizard individual, and r represents an optimal point, which is expressed by the following formula: Among them, q is satisfied The smallest prime number of , i means the variable dimension is the i-th dimension; Mapping the good point set to the feasible region where the population is located, we get: in, represents the position of the kth horned lizard individual in dimension i, ub represents the upper bound of the current dimension, and lb represents the lower bound of the current dimension.

3. The robot optimal path planning method based on the improved horned lizard optimization algorithm according to claim 1 is characterized in that: In step 3, the behavior of horned lizard populations at different stages is improved, specifically including: Add an adaptive dynamic weight q to the hidden behavior of the horned lizard optimization algorithm. The formula is as follows: Among them, maxIdx represents the current worst solution, minIdx represents the current optimal solution, t represents the current number of iterations, Max_iter represents the maximum number of iterations, α and β represent two weight values. As the algorithm iterates, The value gradually increases, The value gradually decreases; Substituting the adaptive dynamic weight q into the hidden behavior formula yields: Among them, x i (t+1) is the position of the horned lizard in the solution search space of the t+1th generation, x best (t) is the optimal solution of generation t; r1, r2, r3 and r4 are integer random numbers generated between 1 and the maximum size of the population, where r1≠r2≠r3≠r4; and is the selected search solution r1, r2, r3, r4; σ is a binary value, Set to 2, c1, c2 are random numbers drawn from Table 29 containing a standardized palette, and c1 ≠ c2.

4. The robot optimal path planning method based on the improved horned lizard optimization algorithm according to claim 1 is characterized in that: The step 3 of improving the behavior of horned lizard populations at different stages also includes: The blood ejection behavior formula of the horned lizard optimization algorithm is modified by replacing Replaced with the dynamic weight ω, the formula is as follows: t represents the current iteration number, Max_iter represents the maximum iteration number; the dynamic weight ω is introduced into the blood ejection behavior formula to obtain: Among them, x i (t+1) is the position of the horned lizard in the solution search space of the t+1th generation, x best (t) is the optimal solution of generation t, x i (t) is the position of the horned lizard in the solution search space of the tth generation, v0 is set to 1, and δ is set to g is the earth's gravity, ε is set to 10 -6 .

5. The robot optimal path planning method based on the improved horned lizard optimization algorithm according to claim 1 is characterized in that: The step 3 of improving the behavior of horned lizard populations at different stages also includes: Modify the escape behavior formula of the horned lizard optimization algorithm and add an adaptive coefficient As shown below: Among them, x i (t+1) is the position of the horned lizard in the solution search space of the t+1th generation, x best (t) is the optimal solution of generation t, x i (t) is the position of the horned lizard in the solution search space of the tth generation, Expressed as walk represents the generation of random numbers between -1 and 1, and γ is the mean of random numbers generated by the standard Cauchy distribution.

6. The robot optimal path planning method based on the improved horned lizard optimization algorithm according to claim 1 is characterized in that: The step 3 of improving the behavior of horned lizard populations at different stages also includes: The skin darkening and lightening behavior formulas of the horned lizard optimization algorithm are modified by adding the Brownian motion random perturbation parameter η; the skin lightening behavior is shown in the following formula: The skin lightening behavior of the horned lizard is shown in the following formula: Where σ is a random binary value. If the value of σ is 1, the skin lightening behavior formula of the horned lizard is executed. If the value of σ is 0, the skin darkening behavior formula of the horned lizard is executed. best (t) is the optimal solution of generation t, Light1 and Light2 are random numbers generated between 0 and 0.4046661, Dark1 and Dark2 are random numbers generated between 0.5440510 and 1, and x worst (t) represents the worst search result value, r1, r2, r3 and r4 are integer random numbers generated between 1 and the maximum number of the population, where r1≠r2≠r3≠r4; and are the selected search solutions r1, r2, r3, and r4.

7. The robot optimal path planning method based on the improved horned lizard optimization algorithm according to claim 1 is characterized in that: The step 3 of improving the behavior of horned lizard populations at different stages also includes: The melanocyte stimulating hormone (MSH) rate formula of the horned lizard optimization algorithm was modified by using a random walk with step size adjustment and an adaptive t-distribution perturbation strategy. The formula is as follows: x r (t)=x best (t)+C_iter(t)·x best (t) Among them, x r (t) is the position of the optimal solution after the adaptive t-distribution mutation disturbance after the t-th iteration of the algorithm; x i (t) is the position of the horned lizard in the solution search space of the tth generation, x best (t) is the optimal solution of generation t, C_iter(t) is the degree of freedom parameter of the distribution with the number of iterations as t, and stepsize represents the step size parameter of the random walk.

Images