AGV path planning and autonomous obstacle avoidance method based on MBESP optimization

Abstract

The application discloses an AGV path planning and autonomous obstacle avoidance method based on MBESP optimization, models an obstacle avoidance environment of an AGV material vehicle, and obtains a model; determines an obstacle avoidance cost function for the model; obtains a vulture search algorithm; improves the obtained vulture search algorithm according to AGV path planning and autonomous obstacle avoidance, and obtains an improved vulture search algorithm; and optimizes AGV path planning and autonomous obstacle avoidance by using the improved vulture search algorithm. The application has better path planning effect when facing different environments, that is, the optimization precision of the improved vulture search algorithm is improved, and an optimal path is found.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of AGV path planning, and particularly relates to an AGV path planning and autonomous obstacle avoidance method based on MBESP optimization. BACKGROUND

[0002] Automated Guided Vehicle (AGV) is widely used in intelligent manufacturing workshops and logistics transportation fields due to its high efficiency, flexible route planning and other advantages. After having a better path planning capability, the AGV can realize full automation in the operation process to a certain extent. However, obstacles often appear in the surrounding environment during the operation process, which will hinder the normal work of the AGV. It is a hot research topic for scholars to plan an optimal path with zero collision from the starting point to each target point while achieving effective autonomous obstacle avoidance. Among them, the improved RRT algorithm proposed for local path planning of the vehicle is complete and fast in search speed, but it usually plans a non-optimal path due to its expansion characteristics. The AGV grid map obstacle avoidance is successfully realized by improving the PSO-PIO algorithm, but it is limited by the combination of the two swarm intelligence methods, and it itself depends on a high number of iterations, so the real-time planning performance is not ideal. A topological grid hybrid modeling method uses genetic algorithm to plan a path, but it has many parameters and slow convergence speed. A relatively smooth shortest path is obtained by using D*Lite, but the global search capability is insufficient in a complex environment. In summary, the optimization effect and iteration convergence speed of the above algorithms for AGV path planning and autonomous obstacle avoidance in a two-dimensional environment all have certain improvement space. The Bald Eagle Search (BES) algorithm is a new meta-heuristic swarm intelligence algorithm proposed by Alsattar in 2020, which simulates the process of bald eagle hunting. The BES algorithm has the characteristics of fast convergence speed, strong adaptability, and easy-to-modify model, and can optimize AGV path planning and autonomous obstacle avoidance. However, the BES algorithm itself has insufficient search capability and defects such as selecting a local optimum, which need to be further improved. SUMMARY

[0003] The present application aims to provide an AGV path planning and autonomous obstacle avoidance method based on MBESP optimization, to solve the problems of complex environment, large path planning calculation amount and other defects in the prior art when AGV plans a path in a two-dimensional space and autonomously avoids obstacles, and the defects of the existing bald eagle search algorithm such as insufficient path search capability and easy to fall into local optimum. The present application has better path planning effect in different environments, i.e. improves the optimization precision of the improved bald eagle search algorithm and finds an optimal path.

[0004] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0005] The AGV path planning and autonomous obstacle avoidance method based on MBESP optimization comprises the following steps:

[0006] Step 1: modeling the AGV material car obstacle avoidance environment, obtaining a model;

[0007] Step 2: determining the obstacle avoidance cost function for the model obtained in step 1;

[0008] Step 3: obtaining a vulture search algorithm;

[0009] Step 4: improving the vulture search algorithm obtained in step 3 according to the AGV path planning and autonomous obstacle avoidance, to obtain an improved vulture search algorithm;

[0010] Step 5: using the improved vulture search algorithm obtained in step 4 to optimize the AGV path planning and autonomous obstacle avoidance.

[0011] Further, the AGV material car obstacle avoidance environment modeling in step 1 is a grid map modeling based on a probability model of the AGV material car, as follows:

[0012]

[0013]

[0014] In the formula, the first formula is an obstacle setting function, and the second formula is a formula for calculating the obstacle grid number after setting the obstacle, m is a map range parameter, n is the number of obstacles, No is the grid number, x k , y k is the coordinate of the top right corner of the unit grid, N is the grid dimension, randperm The function generates a random number between 1 and n; m,n The function arranges n random numbers from small to large, m The function assigns 1 to the position of n , that is, marks the obstacle position as 1. sort G Further, the obstacle avoidance cost function of step 2 is as follows: d d In the grid map, the length of the path and the collision times of the path are weighted and combined to construct a total cost suitable for AGV as follows:

[0015]

[0016]

[0017] ​​​​

[0018] wherein, F s is the cost of the entire path, α is the weight of the path cost, β is the cost weight of the number of collisions, L s is the path length, n p is the number of collisions.

[0019] Further, the step 3 of the vulture search algorithm comprises the following steps:

[0020] Step 3.1: Randomly initialize the vulture population;

[0021] Step 3.2: Select the regional stage position update;

[0022] Step 3.3: Search stage position update;

[0023] Step 3.4: Diving capture stage position update.

[0024] Further, the vulture in step 3.2 identifies and selects the best region to capture prey in the search space, and the selection of the regional stage position update is as follows:

[0025]

[0026] wherein, ɑ is a parameter for controlling the change of the position of the vulture in three-dimensional space, and the value range is (1.5, 2); r is a random number, and the value range is (0, 1); p best is the best search position of the current vulture; p mean is the average value of the two vulture updated positions; p i is the position of the vulture in the population; i

[0027] In step 3.3, the vulture searches for prey in the search space and moves in different directions in the spiral space, gradually accelerating the search range, adopts the spiral flight mode, and updates the position by using the polar equation as follows:

[0028]

[0029]

[0030]

[0031]

[0032] ​

[0033] where, θ ( i ) and r ( i ) are the polar angle and polar radius of the polar coordinate equation, respectively; K and R are the parameters to control the spiral trajectory, which vary in the range of (0, 5) and (0.5, 2), respectively; rand is a random number in (0, 1); x , y are the coordinate values of the bald eagle in polar coordinates, x ( i ) and y ( i ) are the positions of the i th bald eagle in polar coordinates, both of which are in the range of (-1, 1), p i+1 is the next updated position of the i th bald eagle;

[0034] In step 3.4, the bald eagle rapidly dives from the best position in the search space to the target prey, and the other individuals in the population also move to the best position and attack the prey. The movement state is updated using the polar coordinate equation as follows:

[0035]

[0036]

[0037]

[0038]

[0039]

[0040] where, r ( i ) is the polar radius of the polar coordinate equation, and x 1 ，y 1 , xr, yr are the coordinate values of the bald eagle in different cases, x 1 and y 1, xr and yr are the positions of the i th bald eagle in polar coordinates, δ x , δ y are the change values of the coordinate values of the bald eagle in polar coordinates in x and y , respectively, c1 and c 2 represents the inertial parameter of the vulture's movement towards the optimal position, with a value range of (1,2).

[0041] Furthermore, in step 4, the vulture search algorithm obtained in step 3 is improved based on AGV path planning and autonomous obstacle avoidance to obtain an improved vulture search algorithm. The specific steps include:

[0042] Step 4.1: Within the probability-based grid map model obtained in Step 1, randomly initialize the bald eagle population based on the feasible region search of the grid map, with each bald eagle representing an obstacle avoidance path;

[0043] Step 4.2: Update the selection region phase of the bald eagle population using the Cauchy mutation-Levi flight fusion strategy;

[0044] Step 4.3: Update the bald eagle population search phase using a prey guidance mechanism based on a given space;

[0045] Step 4.4: Location update of the bald eagle swarm during the dive-and-capture phase;

[0046] Step 4.5: Update the location of the vulture population using a dynamic adaptive t-distribution mutation operator with feedback mechanism;

[0047] Step 4.6: Calculate the fitness of the bald eagle population and select the best bald eagle based on the fitness value;

[0048] Step 4.7: Determine if the maximum number of iterations has been reached. If yes, output the optimal vulture path, i.e., the optimal obstacle avoidance path. If no, return to step 4.2.

[0049] Furthermore, in step 4.1, the vulture population is randomly initialized based on a feasible region search using a raster map, as shown in the following formula:

[0050]

[0051] In the formula, find ( col The =0) function finds the barrier-free areas in a raster image. length ( find ( col=0 The function finds the length of the barrier-free zone. randi ( length ( find ( col The =0))) function randomly selects a grid cell from the accessible area column. col For a raster map, exclude the start and end columns. ID For the first i The feasible field raster number of the column, X i,d To initialize the bald eagle population;

[0052] Step 4.2, the search space is expanded by using Cauchy variation-Levy flight fusion strategy, and the space stage position update is as follows:

[0053]

[0054]

[0055]

[0056] In the formula, cauchy (0 , 1) refers to the Cauchy sequence subject to (0, 1) distribution, the position parameter is 0 and the scale parameter is 1, Levy Flight is one of the most effective methods to find targets, and its expression is σ is a scaling factor with a variation range of (-1, 1) ，s is Levy Flight path.

[0057] Further, in step 4.3, the Bald Eagle population search stage is updated by using the prey guide mechanism based on the given space, which is specifically:

[0058]

[0059]

[0060] In the formula, p FOOD is the prey position, p fitness is the current Bald Eagle individual fitness value, f FOOD is the current prey fitness value, I is the calculation formula of the position operator.

[0061] Further, step 4.5 uses a dynamic adaptive t-distribution mutation operator with a feedback mechanism to update the position of the Bald Eagle population, which is specifically:

[0062]

[0063]

[0064]

[0065]

[0066]

[0067] In the formula, p i j+1the position of the bald eagle after disturbance, p i j the position of the bald eagle in the first iteration, j the position of the bald eagle in the last iteration, MaxIt the maximum number of iterations, j the current iteration number, w 1 and 2 are respectively the upper limit and the variation range of the dynamic selection probability, w is a sharing coefficient, balancing the size of feedback information in position updating; ζ is a dynamic selection probability operator calculation formula, β is an iteration number parameter, t iter is the optimal bald eagle individual fitness value; F best is a feedback coefficient, showing an increasing trend with the increase of the iteration number, which can accelerate the information exchange between bald eagle individuals; γ i is a random bald eagle individual; b F is the fitness value of the bald eagle individual.

[0068] Further, the improved bald eagle search algorithm obtained in step 5 is used to optimize AGV path planning and autonomous obstacle avoidance, and the specific mode is as follows:

[0069] Each path is a line connected by multiple trajectory points, and a bald eagle in the population is defined as x d i that is, a path, and the bald eagle with the minimum cost represents the optimal path, and each trajectory point has two-dimensional attributes (x, y), when the MBESP is used to optimize AGV path planning and autonomous obstacle avoidance, the specific steps include: x y

[0070] Step 5.1: The initial bald eagle population is obtained by using the method of step 4.1 on the AGV material car environment model obtained in step 1;

[0071] Step 5.2: The fitness of each bald eagle is calculated by using the cost function obtained in step 2 on the initial bald eagle population, the bald eagle population is sorted according to the size of the fitness, and the optimal bald eagle is selected;

[0072] Step 5.3: The position of the bald eagle population is updated by using the methods of steps 4.2-4.5, the fitness of each bald eagle is calculated, and the bald eagle population is sorted according to the size of the fitness;

[0073] Step 5.4: It is judged whether the maximum number of iterations is reached, if yes, the optimal bald eagle, that is, the optimal obstacle avoidance trajectory point is output, otherwise, it is returned to step 5.3;

[0074] ​​​​Step 5.5: the optimal trajectory points obtained in step 5.4 are optimized by adopting an octree search strategy to obtain a final AGV two-dimensional autonomous obstacle avoidance path.

[0075] Compared with the prior art, the present application has the following beneficial technical effects:

[0076] The present application proposes a Multi-Bald Eagle Search based on Prey-guidance (MBESP), which obtains the optimal path of AGV autonomous obstacle avoidance under the premise of ensuring rapid optimization, and has better path planning effect in different environments, i.e., improves the optimization precision of the improved bald eagle search algorithm and finds the optimal path.

[0077] Further, the present application designs a random initialization population based on the feasible region search of the grid map to improve the quality of the initialization population; in the selection of the search space stage, a Cauchy mutation-Levy flight strategy is adopted to solve the problem that BES converges too early in the selection of the search space stage and is easy to fall into a local extreme point; a dynamic adaptive t-distribution mutation operator with a feedback mechanism is designed to improve the global search ability. Thus, MBESP can more accurately and quickly make the AGV material car autonomously avoid the obstacle area and select the optimal path to realize autonomous obstacle avoidance.

[0078] Compared with the bald eagle search algorithm, the present application has better energy consumption, smoother path and faster convergence speed, can effectively make the AGV material car autonomously avoid obstacles, and has better optimization ability. BRIEF DESCRIPTION OF DRAWINGS

[0079] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the present application.

[0080] Figure 1 is a flowchart of the present application;

[0081] Figure 2 is a flowchart of the improved bald eagle search algorithm involved in the present application;

[0082] Figure 3 is a path planning diagram of the present application in a grid map;

[0083] Figure 4 is an effect diagram of the present application in actual application, wherein (a) is an experimental environment, (b) is a completed target, and (c) is a key frame sequence of an experimental video. DETAILED DESCRIPTION

[0084] In order to make the person skilled in the art better understand the technical scheme of the present application, the technical scheme in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the protection scope of the present application.

[0085] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0086] Referring to Figure 1 For the flowchart of the present application, the basic bald eagle search algorithm is optimized and improved, and a multi-strategy fusion bald eagle search algorithm based on prey guidance is proposed, which is applied to AGV path planning and autonomous obstacle avoidance. Figure 2 Firstly, a probability-based grid map model is constructed and a cost function is established. Secondly, a prey guidance mechanism based on a given space is designed to improve the ability of bald eagles to explore and utilize the search space. Thirdly, a dynamic adaptive t-distribution mutation operator with a feedback mechanism is designed to improve the global search ability. Finally, in view of the situation that BES converges too early in the selection space stage and is easy to fall into local extreme points, the Cauchy mutation and Levy flight strategy are introduced to revise and update the formula to jump out of the local optimum. Figure 3 It can be seen that, compared with the bald eagle search algorithm, MBESP has better energy consumption and smoother path, and can make the AGV effectively avoid obstacles autonomously, showing better optimization ability.

[0087] The specific steps are as follows:

[0088] Step 1: Modeling the AGV obstacle avoidance environment includes: AGV based on probability model grid map modeling and multi-objective city point modeling.

[0089] The probability model-based grid map modeling is as follows:

[0090]

[0091]

[0092] In the formula, the first formula is the obstacle setting function, and the second formula is the formula for calculating the obstacle grid number after the obstacle is set. m For map extent parameters, n The number of obstacles. No Number the grid cells. x k , y k The coordinates of the top right corner vertex of the unit grid. N For raster dimensions, randperm ( m,n The function generates 1- m Between n A random number; sort The function sorts these n random numbers in ascending order. G ( d The function )=1 will d The location is assigned a value of 1, meaning the obstacle's location is marked as 1;

[0093] Step 2: The obstacle avoidance cost function for the AGV material vehicle is:

[0094] In the grid map, the costs of path length and number of collisions are weighted and combined to construct a total cost suitable for AGVs as follows:

[0095]

[0096] In the formula, F s The cost of the entire path, α，β Here are the weights of each cost, where α As the weight of the distance cost, β The cost weight for the number of collisions, L s For path length, n p This represents the number of collisions.

[0097] Step 3: The basic vulture search algorithm consists of the following steps:

[0098] The vulture identifies and selects the best area (with the most food) within its search space to capture prey. The search space location is updated as follows:

[0099]

[0100] In the formula, ɑ The parameters used to control the positional changes of the bald eagle in three-dimensional space are set to a value range of (1.5, 2).r This is a random number, with a value range of (0,1). p best This is currently the best search location for vultures; p mean This is an average of the two consecutive bald eagle update locations; p i Let i be the position of the i-th vulture in the population;

[0101] The vulture searches for prey within its search space and moves in different directions within a spiral space, gradually increasing its search range. It employs a spiral flight pattern, and its position is updated using polar coordinate equations as follows:

[0102]

[0103]

[0104]

[0105]

[0106]

[0107] In the formula, θ ( i )and r ( i ) represent the polar angle and polar radius of the polar coordinate equation, respectively; ɑ and R These are parameters that control the spiral trajectory, with variation ranges of (0,5) and (0.5,2), respectively. rand A random number within the range (0,1); x , y These are the coordinates of the vulture in polar coordinates. x ( i )and y ( i ) is the first in polar coordinates i The position of the vulture is given by values ​​(-1, 1). p i+1 For the first i The next update location for the vulture;

[0108] A vulture swoops down from its optimal position in the search space toward its prey, while other individuals in the swarm simultaneously move to their optimal positions and attack. The position is updated using polar coordinates as follows:

[0109]

[0110]

[0111]

[0112]

[0113]

[0114] In the formula, r ( i () represents the polar radius of the polar coordinate equation. To distinguish it from step 3.3, we use... x 1 ，y 1 , xr, yr These are the coordinates of the bald eagle under different conditions. x 1( i )and y 1( i ), xr ( i )and yr ( i ) is the first in polar coordinates i Only the location of the vulture, δ x , δ y The coordinates of the vulture in polar coordinates are respectively... x and y The change value in c 1 and c 2 represents the inertial parameter of the vulture's movement towards the optimal position, with a value range of (1,2).

[0115] Step 3: Improve the vulture search algorithm based on AGV path planning and autonomous obstacle avoidance, including the following steps:

[0116] The raster map obtained in step 1 is used to randomly initialize a bald eagle population based on a feasible region search. Each bald eagle represents a breakthrough path, as shown in the following formula:

[0117]

[0118] In the formula, find ( col The =0) function finds the barrier-free areas in a raster image. length ( find ( col=0 The function finds the length of the barrier-free zone. randi ( length ( find ( col The =0))) function randomly selects a grid cell from the accessible area column. col For a raster map, exclude the start and end columns. ID For the first i The feasible field raster number of the column, X i,dTo initialize the bald eagle population;

[0119] The position update in the search space stage is as follows by using the Cauchy variation-Levy flight fusion strategy:

[0120]

[0121]

[0122]

[0123] In the formula, cauchy (0 , 1) is a Cauchy sequence subject to (0, 1) distribution, the location parameter is 0, and the scale parameter is 1, Levy Flight is one of the most effective methods to find targets, and its expression is σ is a scaling factor with a variation range of (-1, 1) ，s That is, Levy Flight path.

[0124] The bald eagle population search stage is updated by using the prey guide mechanism based on the given space:

[0125]

[0126]

[0127] In the formula, p FOOD is the prey position, p fitness is the current bald eagle individual fitness value, f FOOD is the current prey fitness value, I is the calculation formula of a position operator.

[0128] The formula for the dive capture stage remains unchanged.

[0129] The bald eagle population position is updated by using the dynamic adaptive t-distribution variation operator with a feedback mechanism, which is specifically:

[0130]

[0131]

[0132]

[0133]

[0134]

[0135] In the formula, pi j+1 This indicates the location of the vulture after the disturbance. p i j For the vulture j Position at the next iteration MaxIt The maximum number of iterations, j This is the current iteration number. w 1 and w 2 represents the upper limit and the range of change of the dynamic selection probability, respectively. β It is the sharing coefficient, representing the magnitude of feedback information during equilibrium position updates; ζ is the calculation formula for the dynamic selection probability operator. t ( iter ) represents the number of iterations. F best The optimal fitness value for a single bald eagle; γ The feedback coefficient increases with the number of iterations, which can accelerate information exchange between individual bald eagles. b i These are random individual bald eagles; F This represents the fitness value of an individual bald eagle.

[0136] Calculate the fitness of the new bald eagle population and select the best bald eagle based on the fitness level;

[0137] Determine if the maximum number of iterations has been reached. If so, output the optimal vulture trajectory point, i.e., the optimal obstacle avoidance trajectory.

[0138] Step 5: Optimize AGV path planning and autonomous obstacle avoidance by combining the improved vulture search algorithm. The specific method is as follows:

[0139] Each path is a line connecting multiple trajectory points, defining a vulture in the population. x d i That is, there is a path, and the bald eagle with the lowest cost represents the optimal path. Each trajectory point has two-dimensional attributes ( x , y When using MBESP to optimize AGV path planning and autonomous obstacle avoidance, the specific steps include:

[0140] Step 5.1: Use the method in Step 4.1 to obtain the initial vulture population from the AGV material vehicle environment model obtained in Step 1;

[0141] Step 5.2: Calculate the fitness of each bald eagle in the initial bald eagle population using the cost function obtained in Step 2, sort the bald eagle population according to the fitness, and select the best bald eagle.

[0142] Step 5.3: updating the position of the vulture population by using the method of steps 4.2-4.5, calculating the fitness of each vulture, and ranking the vulture population according to the size of the fitness;

[0143] Step 5.4: determining whether the maximum number of iterations is reached, if yes, outputting the optimal vulture, that is, the optimal obstacle avoidance trajectory point, otherwise, returning to step 5.3;

[0144] Step 5.5: using the octree search strategy to optimize the optimal trajectory point obtained in step 5.4 to obtain the final AGV two-dimensional autonomous obstacle avoidance path.

[0145] Figure 3 For the specific path planning graph in the grid map, Figure 4 For the effect diagram in actual application.

[0146] Figure 3 The path is composed of a feasible region (white region, marked as 0) and an obstacle region (black region, marked as 1), and the comparison between the algorithm MBESP and BES in the AGV autonomous obstacle avoidance route trajectory graph can show that, in the 10x10 grid map experiment, the path planned by BES has a high path cost and a low success rate, and the path cost of MBESP is obviously reduced, and the global path finding ability is improved; in the 20x20 grid map experiment, the search range is further expanded, compared with the optimal path planned by BES, the path searched by MBESP is smoother, the cost is smaller, and the convergence speed is faster. The stability and reliability of MBESP are obviously superior to those of BES.

[0147] Figure 4 In order to verify whether the MBESP proposed in the application can complete the global path planning effect as required in actual application, an AGV developed based on a ROS system is taken as an object for experiment, and the steering diameter is 90cm. The outdoor experimental environment is a section of playground space with obstacles arranged, and the width is 200cm. A laser radar is used to scan the experimental environment and establish a grid map in the ROS system, and the grid is divided according to the principle that the cell width is greater than or equal to the steering radius. Figure 4 (b) is the successful planning of MBESP to reach the target point, Figure 4 (c) is a key frame sequence extracted from the AGV obstacle avoidance experiment video, and the frame number is shown in the upper left corner. At 240 frames, the starting state is started, at 960, 1200, 1440 and 1920 frames, the AGV successfully avoids each obstacle in turn, and at 2160 frames, the AGV reaches the target point and stops working. The experimental results verify the feasibility and effectiveness of MBESP in AGV path planning and autonomous obstacle avoidance.

[0148] It should be pointed out finally that the above embodiments are only used for illustrating the technical solutions of the present application but not for limiting the protection scope thereof, and although the present application has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that after reading the present application, the specific embodiments of the present application can be changed, modified or replaced equivalently by those skilled in the art, but these changes, modifications or equivalent replacements are all within the protection scope of the claims of the present application to be approved.

Claims

1. An AGV path planning and autonomous obstacle avoidance method based on MBESP optimization, characterized in that, Includes the following steps: Step 1: Model the obstacle avoidance environment for the AGV material handling vehicle and obtain the model; Step 2: Determine the obstacle avoidance cost function for the model obtained in Step 1; Step 3: Obtain the vulture search algorithm; Step 4: Improve the vulture search algorithm obtained in Step 3 based on AGV path planning and autonomous obstacle avoidance to obtain the improved vulture search algorithm; specific steps include: Step 4.1: Within the probability-based grid map model obtained in Step 1, randomly initialize the bald eagle population based on grid map feasible region search, with each bald eagle representing an obstacle avoidance path; wherein, the formula for randomly initializing the bald eagle population based on grid map feasible region search is as follows: In the formula, find ( col The =0) function finds the barrier-free areas in a raster image. length ( find ( col=0 The function finds the length of the barrier-free zone. randi ( length ( find ( col The =0))) function randomly selects a grid cell from the accessible area column. col For a raster map, exclude the start and end columns. ID For the first i The feasible field raster number of the column, X i,d To initialize the bald eagle population; Step 4.2: Update the selection area phase of the bald eagle population using the Cauchy mutation-Levi flight fusion strategy; wherein, the Cauchy mutation-Levi flight fusion strategy expands the search space, and the selection space phase position is updated as follows: In the formula, cauchy (0 , 1) refers to a Cauchy sequence that follows a (0,1) distribution, with a position parameter of 0 and a scale parameter of 1. Levy Flight is one of the most effective methods for finding targets, and its expression is as follows: σ It is a scaling factor with a range of (-1, 1). ，s for Levy Flight path, a The parameters used to control the positional changes of the bald eagle in three-dimensional space are set to a value range of (1.5, 2). r This is a random number, with a value range of (0,1). p best This is currently the best search location for vultures; p mean This is an average of the two consecutive bald eagle update locations; p i The first in the population i The location of the vulture; Step 4.3: Update the bald eagle population search phase using a prey guidance mechanism based on a given space; Step 4.4: Location update of the bald eagle swarm during the dive-and-capture phase; Step 4.5: Update the location of the vulture population using a dynamic adaptive t-distribution mutation operator with feedback mechanism; Step 4.6: Calculate the fitness of the bald eagle population and select the best bald eagle based on the fitness value; Step 4.7: Determine if the maximum number of iterations has been reached. If yes, output the optimal vulture path, i.e., the optimal obstacle avoidance path. If no, return to step 4.

2. Step 5: Optimize AGV path planning and autonomous obstacle avoidance using the improved vulture search algorithm obtained in Step 4; MBESP stands for Prey-Guided Multi-Strategy Fusion Vulture Search Algorithm.

2. The AGV path planning and autonomous obstacle avoidance method based on MBESP optimization according to claim 1, characterized in that, In step 1, the obstacle avoidance environment modeling for the AGV material vehicle is a grid map modeling based on a probabilistic model, as shown below: In the formula, the first formula is the obstacle setting function, and the second formula is the formula for calculating the obstacle grid number after the obstacle is set. m For map extent parameters, n The number of obstacles No Number the grid cells. x k , y k The coordinates of the top right corner vertex of the unit grid. N For raster dimensions, randperm ( m,n The function generates 1- m Between n A random number; sort The function sorts n random numbers in ascending order. G ( d The function )=1 will d The location is assigned a value of 1, meaning the obstacle's location is marked as 1.

3. The AGV path planning and autonomous obstacle avoidance method based on MBESP optimization according to claim 2, characterized in that, The obstacle avoidance cost function in step 2 is as follows: In the grid map, the costs of path length and number of collisions are weighted and combined to construct a total cost suitable for AGVs as follows: In the formula, F s The cost of the entire path, α As the weight of the distance cost, β The cost weight for the number of collisions, L s For path length, n p This represents the number of collisions.

4. The AGV path planning and autonomous obstacle avoidance method based on MBESP optimization according to claim 2, characterized in that, Step 3 of the bald eagle search algorithm includes the following steps: Step 3.1: Randomly initialize the vulture population; Step 3.2: Select the area stage location update; Step 3.3: Position update during the search phase; Step 3.4: Position update during the dive capture phase.

5. The AGV path planning and autonomous obstacle avoidance method based on MBESP optimization according to claim 4, characterized in that, In step 3.2, the vulture identifies and selects the best area within the search space to capture prey. The location update for the selected area stage is as follows: In step 3.3, the vulture searches for prey within the search space and moves in different directions within the spiral space, gradually increasing the search range. It employs a spiral flight mode, and its position is updated using polar coordinate equations as follows: In the formula, θ ( i )and r ( i ) represent the polar angle and polar radius of the polar coordinate equation, respectively; K and R These are parameters that control the spiral trajectory, with variation ranges of (0,5) and (0.5,2), respectively. rand A random number within the range (0,1); x , y These are the coordinates of the vulture in polar coordinates. x ( i )and y ( i ) is the first in polar coordinates i The position of the vulture is given by values ​​(-1, 1). p i+1 For the first i The next update location for the vulture; In step 3.4, the vulture swoops down from its optimal position in the search space toward its target prey, while other individuals in the population simultaneously move to their optimal positions and attack the prey. The position is updated using polar coordinate equations as follows: In the formula, r ( i () represents the polar radius of the polar coordinate equation. To distinguish it from step 3.3, we use... x 1 ，y 1 ,xr,yr These are the coordinates of the bald eagle under different conditions. x 1 and y 1, xr and yr In polar coordinates, the first i Only the location of the vulture, δ x δ y The coordinates of the vulture in polar coordinates are respectively... x and y The change value in c 1 and c 2 represents the inertial parameter of the vulture's movement towards the optimal position, with a value range of (1,2).

6. The AGV path planning and autonomous obstacle avoidance method based on MBESP optimization according to claim 5, characterized in that, Step 4.3 employs a prey guidance mechanism based on a predetermined space to update the bald eagle population search phase, specifically as follows: In the formula, p FOOD Location of the prey p fitness This represents the current fitness value of an individual bald eagle. f FOOD This represents the current fitness value of the prey. I Here is the formula for calculating the position operator.

7. The AGV path planning and autonomous obstacle avoidance method based on MBESP optimization according to claim 5, characterized in that, Step 4.5 employs a dynamic adaptive t-distribution mutation operator with feedback mechanism to update the location of the bald eagle population, specifically as follows: In the formula, p i j+1 This indicates the location of the vulture after the disturbance. p i j For the vulture j Position at the next iteration MaxIt The maximum number of iterations, j This is the current iteration number. w 1 and w 2 represents the upper limit and the range of change of the dynamic selection probability, respectively. β It is the sharing coefficient, representing the magnitude of feedback information during equilibrium position updates; ζ is the calculation formula for the dynamic selection probability operator. t ( iter ) represents the number of iterations. F best The optimal fitness value for a single bald eagle; γ The feedback coefficient increases with the number of iterations, which can accelerate information exchange between individual bald eagles. b i These are random individual bald eagles; F This represents the fitness value of an individual bald eagle.

8. The AGV path planning and autonomous obstacle avoidance method based on MBESP optimization according to claim 5, characterized in that, In step 5, the improved vulture search algorithm is used to optimize AGV path planning and autonomous obstacle avoidance. The specific method is as follows: Each path is a line connecting multiple trajectory points, defining a vulture in the population. x d i That is, there is a path, and the bald eagle with the lowest cost represents the optimal path. Each trajectory point has two-dimensional attributes ( x , y When using MBESP to optimize AGV path planning and autonomous obstacle avoidance, the specific steps include: Step 5.1: Use the method in Step 4.1 to obtain the initial vulture population from the AGV material vehicle environment model obtained in Step 1; Step 5.2: Calculate the fitness of each bald eagle in the initial bald eagle population using the cost function obtained in Step 2, sort the bald eagle population according to the fitness, and select the best bald eagle. Step 5.3: Update the position of the vulture population using the methods in steps 4.2-4.5, calculate the fitness of each vulture, and sort the vulture population according to the fitness values; Step 5.4: Determine if the maximum number of iterations has been reached. If yes, output the optimal vulture trajectory point, i.e. the optimal obstacle avoidance trajectory point. Otherwise, return to step 5.

3. Step 5.5: The optimal trajectory points obtained in 5.4 are optimized using an octree search strategy to obtain the final two-dimensional autonomous obstacle avoidance path of the AGV.

Images