A Robot Path Planning Method Based on Knowledge-Learning Artificial Bee Colony Algorithm

By improving the artificial bee colony algorithm, combining environmental modeling and path coding, and adopting a search strategy of hired bees, observer bees, and scout bees, the problem of slow convergence speed and poor performance of the artificial bee colony algorithm in mobile robot path planning is solved, and faster and more efficient global path planning is achieved.

CN117848370BActive Publication Date: 2026-11-17GUIYANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410012783.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-04
Publication Date
2026-11-17
Estimated Expiration
2044-01-04

AI Technical Summary

Technical Problem

The basic artificial bee colony algorithm suffers from slow convergence speed and poor path planning performance in mobile robot path planning, especially in complex environments where its global exploration capability is weak.

Method used

By improving the artificial bee colony algorithm, introducing a knowledge learning mechanism, and combining environmental modeling and path coding, different search strategies such as hired bees, observation bees, and scout bees are adopted to balance global exploration and local mining and optimize path planning.

Benefits of technology

It improves the speed and efficiency of mobile robot path planning, obtains smoother and shorter path planning results, and significantly enhances stability and effectiveness.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117848370B_ABST
    Figure CN117848370B_ABST
Patent Text Reader

Abstract

This invention discloses a robot path planning method based on a knowledge-learning artificial bee colony algorithm. First, the invention models the working environment of the mobile robot, particularly by rationally modeling obstacles within the environment. Second, it establishes a path encoding and optimization model, transforming the mobile robot's path planning problem into a combinatorial constraint optimization problem. Third, it leverages the experiential knowledge gained during algorithm evolution to propose effective learning mechanisms for different search stages, thereby proposing an improved artificial bee colony algorithm. Finally, the improved artificial bee colony algorithm is used to optimize and solve the path planning problem, and simulation experiments are used to verify the feasibility and effectiveness of the proposed method. This invention fully explores and learns effective search experience during the evolution process, proposing corresponding learning operators at different search stages to guide the algorithm towards the region containing the optimal path, thus effectively solving the global path planning problem for mobile robots.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of robotics technology and relates to a robot path planning method based on a knowledge-learning artificial bee colony algorithm, which is used to improve the automation and intelligence of global path planning for mobile robots. Background Technology

[0002] With the development of artificial intelligence technology, mobile robots are widely used in industry, agriculture, service industry, medical rehabilitation and other fields, and people's expectations for their intelligence level are getting higher and higher. Path planning is a core problem that needs to be solved in the application of mobile robots, and it is also one of the key technologies for mobile robots to achieve autonomous navigation. The path planning of mobile robots is essentially a constrained combinatorial optimization problem. Its purpose is to plan an optimal or suboptimal collision-free path from the starting position to the target position for the robot under certain performance metrics. According to the level of intelligence of the path planning method, path planning can be divided into traditional path planning methods and intelligent path planning methods; according to different working environments, mobile robot path planning can be divided into global path planning and local path planning. The intelligence of path planning is an important indicator of the ability of mobile robots to autonomously execute tasks. In recent years, a series of biomimetic intelligent optimization algorithms have been successfully applied to the robot path planning problem, promoting the improvement of the automation and intelligence level of unmanned robot systems, and also broadening the application fields of biomimetic intelligent optimization algorithms.

[0003] Artificial bee colony optimization (ACO) is a relatively new biomimetic swarm intelligence optimization algorithm inspired by the collective intelligent honey-gathering behavior of bees. Due to its simple algorithm structure, few parameters requiring adjustment, ease of implementation, and good optimization performance, ACO has been widely used to solve complex engineering optimization problems in many fields. However, because the optimization equation of the basic ACO has strong randomness, it excels at global exploration but is weak in local exploitation. Consequently, the basic ACO often suffers from slow convergence speed and low convergence accuracy when solving complex optimization problems. Therefore, reasonably balancing exploration and exploitation is key to improving the ACO's ability to solve complex optimization problems. Summary of the Invention

[0004] To address the problems existing in the prior art, the present invention aims to provide a robot path planning method based on a knowledge-learning artificial bee colony algorithm. This invention aims to drive the development of a mobile robot path planning problem by improving the basic artificial bee colony algorithm to propose an effective global path planning method for mobile robots. First, the working environment of the mobile robot is modeled, with particular attention to the reasonable modeling of obstacles. Second, a path encoding and optimization model is established, transforming the mobile robot's path planning problem into a combinatorial constraint optimization problem. Third, based on the empirical knowledge formed during the algorithm's evolution, an effective learning mechanism is proposed for different search stages, leading to an improved artificial bee colony algorithm. Finally, the improved artificial bee colony algorithm is used to optimize and solve the path planning problem, and simulation experiments are used to verify the feasibility and effectiveness of the proposed method.

[0005] In the basic artificial bee colony algorithm, the nectar source location represents a candidate solution to the optimization problem. Both hired bees and observer bees use the same one-dimensional location update equation for the nectar source location. The choice of the update dimension in the equation is random, and the choice of the guiding individual is also random, giving the basic artificial bee colony algorithm strong global exploration capabilities but weak local exploitation capabilities. Therefore, when solving the global path planning problem for mobile robots in complex environments, the basic artificial bee colony algorithm suffers from slow convergence speed and poor path planning performance. This invention proposes a global path planning method for mobile robots based on an improved artificial bee colony algorithm, aiming to effectively balance global exploration and local exploitation in the path planning evolution process of the artificial bee colony algorithm. By fully exploring and learning effective search experience during the evolution process, corresponding learning operators are proposed at different search stages to guide the algorithm to search towards the region where the optimal path is located, thereby effectively solving the global path planning problem for mobile robots.

[0006] The technical solution adopted in this invention is:

[0007] A robot path planning method based on a knowledge-learning artificial bee colony algorithm includes the following steps:

[0008] 1) Perform a two-dimensional model of the mobile robot's working environment to obtain the working environment described in the XOY coordinate system; take the origin of the XOY coordinate system as the starting point of the robot's global planning path and set the target position T; take the line connecting the starting point and the target position T as the horizontal axis X' of the new coordinate system and establish a rotated new coordinate system X'O'Y'; according to the transformation relationship between the XOY coordinate system and the new coordinate system X'O'Y', determine the position coordinates (x, y) in the XOY coordinate system and their corresponding position coordinates (x', y') in the new coordinate system X'O'Y', where θ is the rotation angle of the X'O'Y' coordinate system relative to the XOY coordinate system; divide the line segment OT in the XOY coordinate system into n+1 equal segments, and draw parallel lines l1, l2, l3, ..., l' parallel to the Y' axis through each division point. n Randomly select a path point on each parallel line to generate a candidate path Y' = (y'1, y'2, ..., y') for the robot in the new coordinate system X`O`Y`. n ), y' n The ordinate of the selected path point on the nth parallel line in the candidate path Y';

[0009] 2) Let the size of the nectar source in the artificial bee colony algorithm be SN, and each nectar source i correspond to a candidate path Y. i '=(y' i1 ,y' i2 ,...,y' in );y′ in Candidate path Y i The ordinate of the selected path point on the nth parallel line in the diagram, path Y i The length of ' is L i Let i = 1, 2, ..., SN; the mobile robot path planning problem is transformed into a problem using the formula f(Y) i ')=min(L i +μP i The problem is a minimum optimization problem denoted by ; where min(·) is the minimum function, μ is the constraint violation penalty factor, and P i For path Y i 'Penalty function for collisions with obstacles;'

[0010] 3) The artificial bee colony algorithm is used to solve the minimum optimization problem to obtain the movement path of the mobile robot. The artificial bee colony algorithm includes four stages: initialization, mercenary bee path optimization, observer bee path optimization, and scout bee path optimization. After the initialization stage, the movement path of the mobile robot is obtained through multiple rounds of iterative iteration. Each iteration includes three stages: mercenary bee path optimization, observer bee path optimization, and scout bee path optimization, executed sequentially.

[0011] In the initialization phase, according to y' ij =l' j +rand×(u' j -l' j For path Y i '=(y' i1 ,y' i2 ,...,y' in Initialize; l' j and u' j Let i and j be the lower and upper bounds of the value of the j-th critical path node, respectively, and rand be a random number uniformly distributed in [0,1], where i = 1, 2, ..., SN, j = 1, 2, 3, ..., n;

[0012] In the path optimization stage for hired bees, u is calculated. j =y' ij +r1 j (z j -y' ij ), The intermediate candidate paths U = (u1, u2, ..., u) are obtained. n Then according to For path Y i 'Update according to Update the population-optimal planning path B; j = 1, 2, ..., n; r1 j b is a random number uniformly distributed within the range [-1, 1]. j The j-th critical path point of the optimal planning path B for the population; A path randomly selected in the population The j-th critical path point satisfies k1≠i;

[0013] Observe the bee path optimization process, and randomly select the planned path Y represented by an employed bee i using a roulette wheel strategy. i ',according to Perform a follow-up search to obtain the updated intermediate candidate path U = (u1, u2, ..., u n Then according to For path Y i 'Update according to Update the optimal planning path B for the population; j = 1, 2, ..., n; Let be a random variable uniformly distributed in [0,1], and s be a random number following a Lévy distribution;

[0014] In the scout bee path optimization phase, for a nectar source i that has not evolved for a consecutive limit number of iterations, its corresponding mercenary bee is converted into a scout bee; the scout bee then follows the path planning strategy. The updated planned path Y is obtained. i '; then according to Update the optimal planning path B for the population; j = 1, 2, ..., n; Let be a random variable uniformly distributed in the range [-1, 1]. Let y' be a random variable uniformly distributed within [0,1]; max(y' :j ) represents the maximum value of the j-th dimension component in the current nectar source population; min(y' :j ) represents the minimum value of the j-th dimension component in the current nectar source population.

[0015] Furthermore, the penalty function m is the number of obstacles, P i '(k) represents path Y i 'The collision penalty function with the k-th obstacle in the environment.'

[0016] Furthermore, the touch penalty function v j (k) represents the degree of constraint violation of the j-th path segment by the k-th obstacle, λ j Let be the length of the j-th path segment.

[0017] Furthermore, take q child nodes of the j-th path segment at equal intervals, and examine the distance d between the r-th child node and the center of the k-th obstacle. jkr According to the formula Calculate the degree of violation v of the j-th path segment against the k-th obstacle. j (k); where q is an even number and r is an odd number.

[0018] Furthermore, d is the distance between adjacent equally divided points.

[0019] Furthermore, u satisfies a mean of 0 and a standard deviation of σ. 2 β = 1.5 is a normally distributed random number with mean 0 and standard deviation 1.

[0020] Furthermore, the k-th obstacle in the planning space of the two-dimensional model is... To express; where, (a k ,b k Let r be the center of the k-th obstacle. k Let be the radius of the k-th obstacle.

[0021] Furthermore, the transformation relationship is as follows:

[0022] A robot motion control method is characterized by controlling the robot's motion based on the movement path obtained by the above method.

[0023] A computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps of the above-described method.

[0024] This invention mainly includes the following:

[0025] I. Environmental Modeling

[0026] Ignoring the height information of obstacles, a two-dimensional model of the mobile robot's working environment is constructed. To ensure the robot's safety as it moves along the planned path, obstacles are expanded according to the robot's dimensions and represented by their circumscribed circles. The robot can then be represented as a point mass in the environment. The obstacles in the environment are represented as follows:

[0027]

[0028] In the formula, (x ok ,y ok Let r be the center of the k-th obstacle in the environment. k The radius of the obstacle after expansion.

[0029] Therefore, environmental modeling can be obtained as follows: Figure 1 As shown in the figure, the origin O represents the starting position of the path planning, and the pentagram represents the target position T of the path planning.

[0030] II. Path Planning Optimization Model Building

[0031] In the XOY coordinate system, the line connecting the starting position and the target position T of the path planning is taken as the horizontal axis X' of the new coordinate system, and a coordinate system is established as follows: Figure 2 The new coordinate system X`O`Y` is shown after rotation. The two coordinate systems have the following transformation relationship:

[0032]

[0033] In the formula, x and y are the position coordinates of the robot in the coordinate system XOY, x' and y' are the position coordinates of the robot in the coordinate system X`O`Y`, and θ is the rotation angle of the coordinate system X`O`Y` relative to the coordinate system XOY.

[0034] In the coordinate system X`O`Y`, the line segment O`T` between the starting position and the target position of the path planning is divided into n+1 equal parts. Lines l1, l2, l3, ..., l are drawn parallel to the Y` axis through each division point. n In each parallel line l jIf a path point is randomly selected from (j = 1, 2, ..., n), and this path point is a key node in the planned path, then the n key nodes of the planned path are (x'1, y'1), (x'2, y'2), (x'3, y'3), ..., (x'...). n ,y' n Therefore, a candidate path can be represented by a set of discrete key nodes ((0,0), (x'1,y'1), ..., (x'...). n ,y' n Let (T',0) represent the starting and ending coordinates of the planned path. Here, (0,0) and (T',0) represent the starting and ending coordinates of the planned path, respectively. Since the starting and ending coordinates are known, and the values ​​of the key nodes are restricted to their corresponding parallel lines l... j In this case, the x-coordinates of each node in the planned path in the coordinate system X`O`Y` are known. Therefore, the candidate path can be further simplified to Y'=(y'1,y'2,...,y' n ).

[0035] In the artificial bee colony algorithm, let the size of the nectar source be SN, and each nectar source i correspond to a candidate planning path Y. i '=(y' i1 ,y' i2 ,...,y' in (i = 1, 2, ..., SN), path Y i The length of ' is L i If the optimization objective is to minimize the distance of the planned path, then the path planning problem can be transformed into the following minimum optimization problem.

[0036] f(Y i ')=min(L i +μP i (3)

[0037] In the formula, min(·) is the minimum value function; μ is the constraint violation penalty factor; P i For path Y i 'Penalty function for collisions with obstacles.'

[0038] Path Y i 'Length L i The calculation method is as follows:

[0039]

[0040] In the formula, d is the distance between adjacent equally divided points.

[0041] Penalty function P i The calculation method is as follows:

[0042]

[0043] In the formula, m represents the number of obstacles in the environment; P i '(k) represents path Y i The collision penalty function with the k-th obstacle in the environment is calculated as follows:

[0044]

[0045] In the formula, v j (k) represents the degree of constraint violation of the j-th path segment by the k-th obstacle, λ j Let be the length of the j-th path segment.

[0046] Take q (q is an even number) child nodes from the j-th path segment at equal intervals, and examine the distance d between the odd r-th child node and the center of the k-th obstacle circle. jkr The degree of violation v of the j-th path segment to the k-th obstacle is calculated as follows: j (k).

[0047]

[0048]

[0049] In this invention, if q = 10, then r = 1, 3, 5, 7, 9.

[0050] III. Path Planning Optimization Methods

[0051] In this invention, a candidate planned path for the mobile robot corresponds to the location of a nectar source. Let the size of the nectar source (i.e., the population size of the artificial bee colony algorithm) be SN. Then, in each iteration of the algorithm, SN candidate paths can be obtained, meaning the algorithm can optimize SN candidate paths in parallel. At the end of the algorithm iteration, the optimal planned path obtained is output as the result. The path planning optimization method consists of four stages: initialization, hired bee path optimization, observer bee path optimization, and scout bee path optimization. The optimization method achieves the search for the optimal planned path through the sequential iterative iteration of the hired bee path optimization, observer bee path optimization, and scout bee path optimization stages.

[0052] For path Y i '=(y' i1 ,y' i2 ,...,y' in (i = 1, 2, ..., SN), initialize according to the following formula (9):

[0053] y' ij =l' j +rand×(u' j -l'j (9)

[0054] In the formula, l' j and u' j These are the lower and upper bounds of the values ​​for the j-th (j = 1, 2, 3, ..., n) critical path node, respectively, and rand is a random number uniformly distributed within [0, 1].

[0055] The path planning strategy for hired bees is as follows:

[0056] u j =y' ij +r1 j (z j -y' ij (10)

[0057]

[0058]

[0059]

[0060] In the formula, j = 1, 2, ..., n; r1 j b is a random number uniformly distributed within the range [-1, 1]. j The j-th critical path point of the optimal planning path B for the population; A path randomly selected in the population The j-th critical path point satisfies k1≠i.

[0061] The hired bee search enhances the learning from the global optimal search experience and the different search experiences of other individuals, taking into account both the convergence and diversity of the search. The hired bee search yields intermediate candidate paths U = (u1, u2, ..., u...). n According to the following formula (14), path Y i Update.

[0062]

[0063] The optimal planning path B for the population is updated according to the following formula (15).

[0064]

[0065] After all the hired bees have completed their path search, the observation bees randomly select the planned path Y represented by a hired bee i, based on a roulette wheel selection probability. i Follow the search according to the following formula:

[0066]

[0067] In the formula, j = 1, 2, ..., n; Let be a random variable uniformly distributed within [0,1]; s is a random number following a Lévy distribution, expressed as follows:

[0068]

[0069] In the formula, u is a variable that follows the N(0,σ) rule. 2 ) are normally distributed random numbers, v is a normally distributed random number following N(0,1); in N(x,y), x represents the mean, y represents the variance, and β=1.5.

[0070]

[0071] Observer bee search emphasizes further deep learning from the experience of hired bee search, thereby enhancing the algorithm's development capabilities.

[0072] According to equation (14), the candidate path U = (u1, u2, ..., u3) obtained by the observation bee search is... n For candidate path Y i Update the optimal planning path B for the population according to equation (15).

[0073] For a nectar source i that fails to evolve after a limit number of consecutive iterations, its corresponding mercenary bee is converted into a scout bee. The scout bee's path planning strategy is as follows:

[0074]

[0075] In the formula, j = 1, 2, ..., n; Let be a random variable uniformly distributed in the range [-1, 1]. Let y' be a random variable uniformly distributed within [0,1]; max(y' :j ) represents the maximum value of the j-th dimension component in the current nectar source population; min(y' :j ) represents the minimum value of the j-th dimension component in the current nectar source population.

[0076] Scout bee search takes into account both perturbation learning of existing experience and reverse learning, avoiding the inefficient blind search in the search space in the basic artificial bee colony algorithm.

[0077] The scout bee updates the optimal planned path B for the population according to the following formula.

[0078]

[0079] The advantages of this invention are as follows:

[0080] Based on the basic artificial bee colony algorithm architecture, the method of this invention introduces different experience knowledge learning strategies for hired bees, observation bees, and scout bees. By guiding the evolution process of the algorithm through knowledge, the global exploration and local mining of the algorithm are better balanced, making the global path planning faster and more efficient. Attached Figure Description

[0081] Figure 1 This invention provides a method for modeling the environment for mobile robot path planning.

[0082] Figure 2 The coordinate transformation and path encoding provided in the embodiments of the present invention.

[0083] Figure 3 A path planning flowchart provided for an embodiment of the present invention.

[0084] Figure 4 The mobile robot path planning environment 1 provided in the embodiments of the present invention.

[0085] Figure 5 The mobile robot path planning environment 2 provided in this embodiment of the invention.

[0086] Figure 6 The path planning comparison results are provided in environment 1 of the present invention.

[0087] Figure 7 The path planning comparison results are provided in environment 2 of this embodiment of the invention. Detailed Implementation

[0088] The present invention will now be described in further detail with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0089] The core idea of ​​this invention is to perform two-dimensional modeling of the working environment of the mobile robot and use key nodes for path encoding; to transform the path planning problem of the mobile robot into a minimum combinatorial optimization problem; to build a learning mechanism of empirical knowledge and improve the basic artificial bee colony algorithm, thereby achieving efficient optimization of the global path planning of the mobile robot.

[0090] The detailed execution steps of the global path planning method in this embodiment of the invention are as follows:

[0091]

[0092]

[0093] use Figure 4 and Figure 5The effectiveness of the path planning method of this invention was tested in two environments with different levels of complexity. The method of this invention was compared experimentally with a path planning method based on the basic artificial bee colony algorithm. The relevant parameters were set as follows: the bee colony size for all algorithms was 40, the maximum number of iterations was 500, the limit parameter of the algorithm was 100, and the penalty factor μ was 100.

[0094] The simulation experiment and its results are as follows:

[0095] In two different environments, path optimization was performed 10 times consecutively using two different methods. The path planning results are as follows: Figure 6 and Figure 7 As shown in the figure, the dashed path is obtained by the path planning method based on the basic artificial bee colony algorithm, and the solid path is obtained by the planning method of this invention.

[0096] according to Figure 6 Compared with the comparison method, the planning method of the present invention obtained similar planning results for each path planning in environment 1, while the results of the comparison method varied greatly for each path planning; the path obtained by the method of the present invention is smoother and shorter in each path planning.

[0097] according to Figure 7 Compared with the comparative method, the planning method of the present invention achieves similar planning results for each path planning in complex environment 2, while the results of each path planning of the comparative method vary greatly and a more serious "return" phenomenon occurs; the path planning method of the present invention obtains smoother and shorter paths for each path planning.

[0098] In summary, compared with the comparative path planning methods, the path planning method of the present invention has a stronger global search capability, and its path planning results are more stable and the planning effect is better.

[0099] Although specific embodiments of the invention have been disclosed for illustrative purposes to aid in understanding and implementing the invention, those skilled in the art will understand that various substitutions, variations, and modifications are possible without departing from the spirit and scope of the invention and the appended claims. Therefore, the invention should not be limited to the content disclosed in the preferred embodiments, and the scope of protection claimed by the invention is defined by the claims.

Claims

1. A robot path planning method based on a knowledge-learning artificial bee colony algorithm, comprising the following steps: 1) Perform a two-dimensional model of the working environment of the mobile robot to obtain the working environment described in the XOY coordinate system; The origin of the XOY coordinate system is used as the starting point of the robot's global planning path, and the target position T is set. A new, rotated coordinate system X`O`Y` is established with the line connecting the starting point and the target position T as the horizontal axis X` of the new coordinate system. Based on the transformation relationship between the XOY coordinate system and the new X`O`Y`, the corresponding position coordinates (x', y') in the new coordinate system X`O`Y` are determined, where θ is the rotation angle of the X`O`Y` coordinate system relative to the XOY coordinate system. The line segment OT in the XOY coordinate system is divided into n+1 equal segments, and parallel lines l1, l2, l3, ..., l'T' are drawn through each division point parallel to the Y` axis. n Randomly select a path point on each parallel line to generate a candidate path Y' = (y′1, y′2, ..., y′) for the robot in the new coordinate system X`O`Y`. n ), y′ n The ordinate of the selected path point on the nth parallel line in the candidate path Y'; 2) Let the size of the nectar source in the artificial bee colony algorithm be SN, and each nectar source i correspond to a candidate path Y. i ′=(y′ i1 ,y′ i2 ,...,y′ in );y′ in Candidate path Y i The ordinate of the selected path point on the nth parallel line in the diagram, path Y i The length of ′ is L i Let i = 1, 2, ..., SN; the mobile robot path planning problem is transformed into a problem using the formula f(Y) i ′)=min(L i +μP i The problem is a minimum optimization problem denoted by ; where min(·) is the minimum function, μ is the constraint violation penalty factor, and P i For path Y i The penalty function for 'colliding with an obstacle'; 3) The artificial bee colony algorithm is used to solve the minimum optimization problem to obtain the movement path of the mobile robot. The artificial bee colony algorithm includes four stages: initialization, mercenary bee path optimization, observer bee path optimization, and scout bee path optimization. After the initialization stage, the movement path of the mobile robot is obtained through multiple rounds of iterative iteration. Each iteration includes three stages: mercenary bee path optimization, observer bee path optimization, and scout bee path optimization, executed sequentially. In the initialization phase, according to y′ ij =l′ j +rand×(u′ j -l′ j For path Y i ′=(y′ i1 ,y′ i2 ,...,y′ in Initialize; l′ j and u′ j Let i and j be the lower and upper bounds of the value of the j-th critical path node, respectively, and rand be a random number uniformly distributed in [0,1], where i = 1, 2, ..., SN, j = 1, 2, 3, ..., n; In the path optimization stage for hired bees, u is calculated. j =y′ ij +r1 j (z j -y′ ij ), The intermediate candidate paths U = (u1, u2, ..., u) are obtained. n Then according to For path Y i Update according to Update the population-optimal planning path B; j = 1, 2, ..., n; r1 j b is a random number uniformly distributed within the range [-1, 1]. j The j-th critical path point of the optimal planning path B for the population; A path randomly selected in the population The j-th critical path point satisfies k1≠i; Observe the bee path optimization process, and randomly select the planned path Y represented by an employed bee i using a roulette wheel strategy. i ',according to Perform a follow-up search to obtain the updated intermediate candidate path U = (u1, u2, ..., u n Then according to For path Y i Update according to Update the optimal planning path B for the population; j = 1, 2, ..., n; Let be a random variable uniformly distributed in [0,1], and s be a random number following a Lévy distribution; In the scout bee path optimization phase, for a nectar source i that has not evolved for a consecutive limit number of iterations, its corresponding mercenary bee is converted into a scout bee; the scout bee then follows the path planning strategy. The updated planned path Y is obtained. i ′; then according to Update the optimal planning path B for the population; j = 1, 2, ..., n; Let be a random variable uniformly distributed in the range [-1, 1]. Let y be a random variable uniformly distributed in [0,1]; max(y′) :j ) represents the maximum value of the j-th dimension component in the current nectar source population; min(y′) :j ) represents the minimum value of the j-th dimension component in the current nectar source population.

2. The method according to claim 1, characterized in that, penalty function m is the number of obstacles, P i ′(k) represents path Y i The collision penalty function with the k-th obstacle in the environment.

3. The method according to claim 2, characterized in that, Touch penalty function v j (k) represents the degree of constraint violation of the j-th path segment by the k-th obstacle, λ j Let be the length of the j-th path segment.

4. The method according to claim 3, characterized in that, Take q child nodes of the j-th path segment at equal intervals, and examine the distance d between the r-th child node and the center of the k-th obstacle circle. jkr According to the formula Calculate the degree of violation v of the j-th path segment against the k-th obstacle. j (k); where q is an even number and r is an odd number.

5. The method according to claim 1, characterized in that, d is the distance between adjacent equally divided points.

6. The method according to claim 1, characterized in that, u satisfies a mean of 0 and a standard deviation of σ. 2 β = 1.5 is a normally distributed random number with mean 0 and standard deviation 1.

7. The method according to claim 1, characterized in that, The k-th obstacle in the planning space of the two-dimensional model is adopted. To express; Where (ak, bk) is the center of the k-th obstacle, r k Let be the radius of the k-th obstacle.

8. The method according to claim 1, characterized in that, The transformation relationship is as follows:

9. A robot motion control method, characterized in that, The robot's motion is controlled based on the movement path obtained by the method described in claim 1.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.