Inspection robot multi-target-point path planning method based on improved dung beetle optimization algorithm

Abstract

The invention discloses an inspection robot multi-target-point path planning method based on an improved dung beetle optimization algorithm, and relates to the field of path planning. The method comprises the following steps: constructing a grid map of a to-be-planned area, and marking a path starting point, a target point, a free grid and an obstacle grid on the grid map; taking the population generated according to the grid map as an initial dung beetle population in dung beetle calculation; after position updating is carried out on the dung beetle group according to the barrier-free condition and the barrier condition, a first optimal solution of the dung beetle group is solved according to the fitness function; after the breeding dung beetle group, the foraging dung beetle group and the stealing dung beetle group are subjected to position updating, a second optimal solution is obtained through the fitness function, and the first optimal solution and the second optimal solution are combined to obtain a new population; and repeating the above steps until the maximum number of iterations is reached, and finally removing redundant nodes and carrying out second-order Bezier curve smoothing processing to obtain an optimal path.

Description

Technical Field

[0001] The invention relates to the field of path planning, and in particular to a multi-objective path planning method for an inspection robot based on an improved dung beetle optimization algorithm. Background Art

[0002] The path planning of the inspection robot can be simply described as a TSP problem. The goal of TSP is to find the shortest access route that meets the above conditions. It is difficult to obtain the optimal solution using general algorithms. In the early days, precise algorithms such as branch and bound and dynamic programming were used to solve it. However, as the scale of the problem is expanded, it will be difficult to obtain accurate results, so heuristic algorithms are proposed. Although the results obtained by the heuristic algorithm cannot be proved to be the optimal solution, it can effectively provide suboptimal or possibly optimal solutions. It is divided into traditional heuristic algorithms and meta-heuristic algorithms.

[0003] Dung Beetle Optimizer (DBO) is a new type of swarm intelligence optimization algorithm. Compared with classic algorithms such as particle swarm optimization and genetic algorithm, DBO algorithm takes into account global exploration and local development, and has the characteristics of fast convergence speed and high solution accuracy. DBO algorithm divides different survival tasks according to the division of labor of dung beetle populations. The population is divided into four types: rolling dung beetles, breeding dung beetles, foraging dung beetles, and stealing dung beetles. However, the no free lunch theorem logically proves that no metaheuristic algorithm is suitable for solving all optimization problems. In terms of optimization problems, DBO algorithm is superior to other algorithms. However, it is still very challenging for DBO algorithm to obtain the ideal optimal solution. In addition, its ability to solve complex problems is not satisfactory. Although DBO algorithm has the characteristics of strong optimization ability and fast convergence speed, it also has the disadvantages of imbalance between global exploration and local development ability, easy to fall into local optimality, and weak global exploration ability. Summary of the invention

[0004] In view of the above-mentioned defects of the prior art, the present invention provides a multi-target point path planning method for an inspection robot based on an improved dung beetle optimization algorithm, comprising the following steps:

[0005] S1. Construct a grid map of the area to be planned, and mark the path starting point, target point, free grid and obstacle grid on the grid map;

[0006] S2, generating a population P = {p1, p2, ... p n} is used as the initial population of dung beetles in the dung beetle calculation;

[0007] S3, dividing the population into a rolling dung beetle group, a breeding dung beetle group, a foraging dung beetle group and a thieving dung beetle group; after the rolling dung beetle group updates its position according to the barrier-free situation and the barrier situation, a first optimal solution of the rolling dung beetle group is obtained according to the fitness function; after the breeding dung beetle group, the foraging dung beetle group and the thieving dung beetle group respectively update their positions, a second optimal solution is obtained through the fitness function, and the first optimal solution and the second optimal solution are combined to obtain a new population;

[0008] S4, repeat the above step S3 until the maximum number of iterations is reached, and the current optimal path is obtained;

[0009] S5, removing redundant nodes from the current optimal path to obtain a new current optimal path;

[0010] S6. Smoothing the new current optimal path described in step S5 to obtain a final optimal path.

[0011] Furthermore, the initial population is initialized by Logistic chaotic mapping, and the Logistic chaotic mapping is:

[0012] x i+1 =ax i (1-x i );

[0013] Among them, a is the control parameter.

[0014] Furthermore, the method for updating the position of the rolling ball dung beetle is:

[0015]

[0016] Among them: R1, R2 are random numbers, R1∈[0,2π], R2∈[0,π], r1, r2 are golden section coefficients, which are used to narrow the search space and guide the individual to converge to the optimal value. The segmentation is achieved as follows:

[0017] r1=u(1-τ)+vτ

[0018] r2=uτ+v(1-τ)

[0019] Among them, u and v are the initial values ​​of the golden section coefficients, u = π, v = -π,

[0020] Furthermore, the position update process of the dancing dung beetle is as follows:

[0021]

[0022] Among them: θ≠0, π.

[0023] Furthermore, the method for updating the position of the foraging dung beetle is:

[0024]

[0025] Furthermore, the method for updating the position of the stealing dung beetle is:

[0026]

[0027] Among them, g represents a random vector of size 1×D, which obeys the normal distribution, and S represents a constant value.

[0028] Compared with the prior art, the present invention has the following technical effects:

[0029] The present invention improves the dung beetle optimization algorithm under the background of multi-target point path planning. First, the dung beetle optimization algorithm calls the Logistic chaotic mapping to change its initialization process, enhance the diversity of the population, and accelerate the convergence speed. Secondly, the rolling behavior and foraging behavior determine the algorithm's exploration ability and convergence speed of the solution space, so the Gold-SA algorithm is introduced to improve the dung beetle rolling behavior and foraging behavior position update, effectively coordinate the global search ability and local mining ability, and accelerate the convergence speed. Finally, the Levy flight strategy is introduced into the dung beetle stealing behavior position update formula to enhance the algorithm's ability to jump out of the local optimum. Finally, the redundant nodes are eliminated, and the global path is smoothed by a second-order Bezier curve to obtain the optimized path.

[0030] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 is a schematic flow chart of a path planning method in a specific embodiment of the present invention;

[0032] Figure 2 It is a schematic diagram of the distribution of a chaotic sequence of a Logistic map according to a specific embodiment of the present invention;

[0033] Figure 3 is a schematic diagram of the proportions of various dung beetle groups according to a specific embodiment of the present invention;

[0034] Figure 4 is a schematic diagram of a conceptual model of a boundary selection strategy in a specific embodiment of the present invention;

[0035] Figure 5 It is a position change diagram of the Levy flight strategy in a two-dimensional coordinate system according to a specific embodiment of the present invention;

[0036] Figure 6It is a schematic diagram of path smoothing according to a specific embodiment of the present invention. DETAILED DESCRIPTION

[0037] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0038] While some exemplary embodiments of the present invention have been described for the purpose of illustration, it should be understood that the present invention may be implemented in other ways not specifically shown in the drawings.

[0039] like Figure 1 As shown, in a specific embodiment, a multi-target point path planning method for an inspection robot based on an improved dung beetle optimization algorithm is provided, comprising the following steps:

[0040] S1. Construct a grid map of the area to be planned, and mark the path starting point, the coordinates of all target points, the free grid and the obstacle grid on the grid map, set the width of the grid map, the scale labels of the two coordinate axes, set the traversal path array for drawing the line segments connecting the points, and set the line width.

[0041] In this embodiment, the first grid in the lower left corner is taken as the starting point, the lower left corner of each grid is taken as the punctuation point, the starting point is (0, 0), and the grid to the right of the starting point is (0, 1). Multiple target points are set. To avoid collision between the robot and obstacles, the obstacles are expanded, and finally the freely traversable areas and inaccessible areas in the map are determined. The grid type is filled with white for free grids and gray for obstacle grids.

[0042] S2, generating a population P = {p1, p2, ... p n} is the initial population of dung beetles in the dung beetle calculation, and each dung beetle represents a possible path;

[0043] S3, dividing the population into a rolling dung beetle group, a breeding dung beetle group, a foraging dung beetle group and a thieving dung beetle group; after the rolling dung beetle group updates its position according to the barrier-free situation and the barrier situation, the first optimal solution of the rolling dung beetle group is obtained according to the fitness function; after the breeding dung beetle group, the foraging dung beetle group and the thieving dung beetle group update their positions respectively, the second optimal solution is obtained through the fitness function, and the first optimal solution and the second optimal solution are combined to obtain a new population.

[0044] S31. Divide the dung beetle population into a rolling dung beetle group, a breeding dung beetle group, a foraging dung beetle group and a stealing dung beetle group.

[0045] The dung beetle optimization algorithm is a heuristic optimization algorithm that performs global search by simulating the rolling, foraging, stealing and reproduction behaviors of dung beetles. It can find the optimal path in the entire search space. The dung beetle algorithm has strong global search capabilities, fast convergence speed and high solution accuracy, which can help robots find the global optimal solution in complex environments.

[0046] In the prior art, each initial group is divided by random allocation, which may lead to problems such as low group diversity, uneven group distribution, and rapid convergence to the local optimal solution. This embodiment uses Logistic chaotic mapping as a method to improve the diversity of the population initialized by the dung beetle algorithm to obtain a more uniform and diverse initial group, thereby improving the solution accuracy and convergence speed of the intelligent algorithm. Figure 2 The Logistic chaotic map shown is:

[0047] x i+1 =ax i (1-x i );

[0048] Wherein, a is a control parameter, and its value is in (0,4]. The larger a is, the higher the chaos is. When a=4, it is in a completely chaotic state, and the chaotic orbit state value range is (0,1). In this embodiment, a is 3.99, and the initial value x=0.66.

[0049] After the Logistic chaotic mapping process, the population parameters are initialized, including the population size pop, the maximum number of iterations Max, the dimension N, the upper boundary UB and the lower boundary LB. In this embodiment, the proportion of the rolling dung beetle group is 0.2, the proportion of the egg-laying dung beetle group is 0.2, the proportion of the small dung beetle group is 0.25, and the proportion of the stealing dung beetle group is 0.35. The ratio of dung beetles in each group is, for example, Figure 3 shown.

[0050] S32. Set the fitness function.

[0051] Determine whether the individual path of each dung beetle passes through an obstacle. If there is no obstacle, a continuous and smooth path is generated. The fitness value is determined by the Euclidean distance between all adjacent points in the path; if there is an obstacle in the path, the fitness value is set to a number to penalize the obstacle in the path. The calculation method is to multiply the number of obstacles by the total area of ​​the map (that is, the number of rows and columns of the map), thereby reflecting the undesirability of the path. The fitness function is:

[0052]

[0053] Where nB represents the number of times the path passes through obstacles, x i,1 and x i,2 are the x-axis coordinates and y-axis coordinates of the ith point in the path, N is the total number of points on the path, and E(X) and E(Y) represent the number of rows and columns of the map.

[0054] S33. A group of ball-rolling dung beetles performs ball-rolling behavior.

[0055] Dung beetles have an interesting habit of making a ball of feces and then rolling it to the desired location. During the rolling process, the dung beetle needs to use celestial clues (sun position or wind direction, etc.) to keep the dung ball rolling in a straight line. In order to simulate the rolling behavior, the dung beetle needs to move in a given direction throughout the search space. In this mode, the position of the dung beetle during the rolling process is updated as follows:

[0056]

[0057] Where: k∈(0,0.2] represents the deflection coefficient, b is a constant between (0,1), t represents the number of iterations, represents the position of the dung beetle, a is a natural coefficient assigned a value of -1 or 1, 1 means no deviation, and -1 means deviation from the original direction. represents the worst position in the current population, Used to simulate changes in light intensity.

[0058] In this embodiment, the position update of the rolling ball behavior is improved by the golden sine algorithm (Gold-SA):

[0059]

[0060] Among them: R1, R2 are random numbers, R1∈[0,2π], R2∈[0,π], r1, r2 are golden section coefficients, which are used to narrow the search space and guide the individual to converge to the optimal value. The segmentation is achieved as follows:

[0061]

[0062] Among them, u and v are the initial values ​​of the golden section coefficients, u = π, v = -π,

[0063] When the dung beetle encounters an obstacle and cannot move forward, it needs to adjust its direction by dancing to obtain a new route. This embodiment uses a tangent function to simulate the dancing behavior of the dung beetle to obtain a new rolling direction. Once the dung beetle successfully determines a new direction, it will continue to roll the ball forward, and the position update at this time is as follows:

[0064]

[0065] Among them, when θ=0, π / 2 or π, tan(θ) is 0 or has no meaning, so the position of the dung beetle will not be updated when θ=0, π / 2 or π.

[0066] S34. The breeding dung beetle group performs breeding behavior, the foraging dung beetle group performs foraging behavior, and the thieving dung beetle group performs stealing behavior.

[0067] In nature, dung balls are rolled to a safe place and hidden by dung beetles. In order to provide a safe environment for their offspring, it is crucial for dung beetles to choose a suitable place to lay eggs. In this embodiment, the boundary selection strategy is used to simulate the area where female dung beetles lay eggs as follows:

[0068]

[0069] Where: R = 1-t / T, T is the maximum number of iterations, Lb and Ub are the lower and upper bounds respectively, is the global optimal position of the current population, Lb * and Ub * They are the lower and upper bounds of the egg-laying area respectively, which means that the area where dung beetles lay eggs is dynamically adjusted with the number of iterations.

[0070] like Figure 4 As shown, the blue dot represents the position of the rolling dung beetle, the current local best position It is represented by a large orange circle, and The surrounding black circles represent breeding balls. Each breeding ball contains a dung beetle egg, and the red circles represent the upper and lower boundaries.

[0071] Some dung beetles that have grown into adults crawl out of the ground to look for food. They are called small dung beetles. In addition, it is necessary to establish an optimal foraging area to guide the dung beetles to forage, which simulates the foraging process of these dung beetles in nature. This behavior is mainly for small dung beetles. The optimal foraging area of ​​small dung beetles is dynamically updated, as shown below:

[0072]

[0073] in: is the local optimal position of the current population. l and Ub l Define the lower and upper bounds of the dung beetle's foraging area. The position of the dung beetle is updated as follows:

[0074]

[0075] Where: C1 is a random number that obeys the normal distribution, and C2 represents a random vector between (0,1).

[0076] In this embodiment, for foraging behavior, the obtained golden ratio is used for position update, and the position update method of the foraging dung beetle group is:

[0077]

[0078] in: is the optimal position of the i-th individual at the t-th iteration; r1 and r2 are random numbers, and the segmentation is implemented as follows:

[0079]

[0080] Among them: r1∈[0,2π], r2∈[0,π], a, b are the initial values ​​of the golden section, the golden sine is divided by the standard sine interval, since the period of the sine function sin(x) is 2π, according to its definition and the relationship with the unit element, in order to enable the population to traverse the searched space of each dimension in the entire period, here a=π, b=-π. τ is the golden section ratio, usually taken The golden ratio coefficient reduces the space with each iteration, and each division guides the individual to gradually approach the optimal value.

[0081] Some dung beetles, called thieves, steal dung balls from other dung beetles. The position update formula for the thieves is:

[0082]

[0083] Where: g represents a random vector of size 1×D, which obeys the normal distribution, and S represents a constant value.

[0084] From the above formula, we can see that is the best food source. Therefore, it can be assumed that Nearby is the best place to compete for food. During the iteration process, the location information of the stealing dung beetle is updated as follows:

[0085] Levy flight is a type of random walk strategy whose moving step length conforms to a heavy-tailed stable distribution, which is manifested as alternating movements of shorter distances and relatively longer distances. The position change diagram in a two-dimensional coordinate system is as follows: Figure 4As shown, the horizontal axis is the x-axis and the vertical axis is the y-axis. The Levy flight strategy is used to improve the randomness of the individual search step, improve the effectiveness of the population search range and the global comprehensive search capability, ensure the information supply of excellent individuals, and make the algorithm's movement mode alternate between frequent short-distance jumps and occasional long-distance jumps, so as to facilitate jumping out of the local optimum. The use of the Levy flight search mechanism in the bionic swarm intelligent optimization algorithm can effectively increase the exploration range of the search space, improve the diversity of group search, and help the search algorithm jump out of the local optimum. The optimized formula is:

[0086]

[0087] During the stealing behavior stage, there may be problems such as insufficient search range, falling into local optimal solutions and convergence failure. Therefore, the Levy strategy is introduced into the position update formula of the thief dung beetle to enhance the algorithm's ability to jump out of the local optimal solution.

[0088] After each of the above actions, it is necessary to determine whether the updated position information exceeds the boundary value, change the value that exceeds the boundary value to the closest boundary value, and then update the fitness value fitness and the optimal solution; through the dung beetle algorithm, a collision-free path from the starting point to the target point is accurately and effectively found, which solves the problem of falling into local optimality and uncertain results of previous swarm intelligence algorithms. Multiple cycles are used to determine whether the fitness value is the current optimal fitness value. If it is, it is updated and determined whether it is the global optimal, and finally the global optimal fitness value is updated to the individual optimal fitness. The fitness of the optimal solution of each generation is recorded to track the efficiency and effectiveness of the optimization process.

[0089] S4. Repeat step S3 above until the maximum number of iterations is reached to obtain the current optimal path.

[0090] S5. Eliminate redundant nodes from the current optimal path. Starting from the starting point (first node) of the current optimal path, check each node in sequence. If a node is found to be directly connected to the subsequent node on the path through a straight line, and the straight line does not pass through other nodes on the path, remove the node and update the connection mode of its adjacent nodes until all nodes are traversed to obtain a new current optimal path. In this embodiment, for any node i on the path, the following redundant node elimination measures are taken:

[0091] S51. Initialize nodes: define each node in the grid map. These nodes may be intersections, crossroads or other important geographical locations on the map.

[0092] S52, obstacle detection: For the connection between the current node i and the subsequent node, when an obstacle appears on the straight line segment between i and j (where j>i), try to directly connect i and j-1 with a straight line segment.

[0093] S53, node deletion: If it is found that nodes i and j-1 can be directly connected, delete all intermediate nodes between i and j-1.

[0094] S54. Update path: connect node i directly to node j-1.

[0095] S55, repeated detection: Continue to perform the same detection on the updated path until all nodes are evaluated. To ensure optimization, multiple iterations can be performed until no further nodes can be deleted and the optimal path is obtained.

[0096] S6. Smoothing the optimal path obtained in step S5.

[0097] When the path is not smooth enough, the robot will generate unnecessary friction loss, and the non-smooth path will also increase the time loss of the robot when turning, reduce efficiency, and redundant turning points will increase the path length. In this embodiment, the optimal path obtained by using the second-order Bezier curve to improve step S5 has many turning points and is not smooth. The second-order Bezier curve is as follows: Figure 6 shown.

[0098] In the figure, p0, p2 is a parabola passing through three different points; the two tangent lines passing through points p0 and p2 intersect at point p1. The tangent line of point intersects p2p1 and p2p1 at and The following formula is established:

[0099]

[0100] The current optimal path obtained by S5 is optimized by the second-order Bezier curve using the following formula:

[0101] P(t)=(1-t)p0+2t(1-t)p1

[0102] The final optimal path is obtained by smoothing the global path.

[0103] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A multi-target point path planning method for an inspection robot based on an improved dung beetle optimization algorithm, characterized in that: The steps include: S1. Construct a grid map of the area to be planned, and mark the path starting point, target point, free grid and obstacle grid on the grid map; S2, generating a population P = {p1, p2, ... p n } is used as the initial population of dung beetles in the dung beetle calculation; S3, dividing the population into a rolling dung beetle group, a breeding dung beetle group, a foraging dung beetle group and a thieving dung beetle group; after the rolling dung beetle group updates its position according to the barrier-free situation and the barrier situation, a first optimal solution of the rolling dung beetle group is obtained according to the fitness function; after the breeding dung beetle group, the foraging dung beetle group and the thieving dung beetle group respectively update their positions, a second optimal solution is obtained through the fitness function, and the first optimal solution and the second optimal solution are combined to obtain a new population; S4, repeat the above step S3 until the maximum number of iterations is reached, remove the redundant nodes of the path, and optimize the path by a second-order Bezier curve to obtain the current optimal path; S5, removing redundant nodes from the current optimal path to obtain a new current optimal path; S6. Smoothing the new current optimal path described in step S5 to obtain a final optimal path.

2. According to claim 1, a multi-target point path planning method for an inspection robot based on an improved dung beetle optimization algorithm is characterized in that: The initial population is initialized by Logistic chaotic mapping, and the Logistic chaotic mapping is: x i+1 =ax i (1-x i ); Among them, a is the control parameter.

3. The multi-target point path planning method of the inspection robot based on the improved dung beetle optimization algorithm according to claim 1 is characterized in that: The method for updating the position of the rolling ball dung beetle is as follows: Among them: R1, R2 are random numbers, R1∈[0,2π], R2∈[0,π], r1, r2 are golden section coefficients, which are used to narrow the search space and guide the individual to converge to the optimal value. The segmentation is achieved as follows: r1=u(1-τ)+vτ r2=uτ+v(1-τ) Among them, u and v are the initial values ​​of the golden section coefficients, u = π, v = -π, 4. The inspection robot multi-target point path planning method based on the improved dung beetle optimization algorithm according to claim 1 is characterized in that: The position update process of the dancing dung beetle is as follows:

5. The inspection robot multi-target point path planning method based on the improved dung beetle optimization algorithm according to claim 1 is characterized in that: The method for updating the position of the foraging dung beetle is as follows:

6. The inspection robot multi-target point path planning method based on the improved dung beetle optimization algorithm according to claim 1 is characterized in that: The method for updating the position of the stealing dung beetle is as follows: Among them, g represents a random vector of size 1×D, which obeys the normal distribution, and S is a constant value.

Images