An underwater robot multi-task point path planning method considering the marine environment

Abstract

The present invention discloses a multi-task point path planning method for an underwater robot considering the marine environment. First, a map is constructed according to the grid method, and a current function is constructed according to the current situation; then, an evaluation system for path planning is determined, and a fitness function is constructed; then, the A* algorithm is used to find the shortest distance between any two points, and the distance matrix is transmitted to the improved ant colony algorithm; the relevant parameters of the ant algorithm are initialized, the number of ants and the maximum number of iterations are set; then, the improved ant colony algorithm is carried out by using an adaptive pheromone factor, heuristic function factor, and evaporation factor; then, the roulette method is used to move the position of the ants until the target point is reached, and the pheromone is updated after all ants complete one iteration process; the iteration process is repeated until all ants reach the maximum number of iterations, and the optimal path planning result is output. The improved ant colony algorithm proposed by the present invention reduces the convergence time and obtains different optimization results according to different requirements.

Description

Technical Field

[0001] The present invention relates to a path planning method for an underwater robot, and particularly to a multi-task point path planning method for an underwater robot considering the ocean environment. Background Art

[0002] With the development of science and technology, underwater robots play an important role in fields such as ocean exploration, resource exploration, and environmental monitoring. Among them, the path planning technology of underwater robots is particularly crucial, which enables the robots to efficiently complete the detection, observation, and sampling work at multiple task points in the ocean environment. The main purpose of path planning is to find one or more optimal driving routes given a starting point and an ending point. These routes need to meet the requirements of being safe, effective, and as short as possible or conforming to other performance indicators (such as time, energy consumption, smoothness, etc.). The successful application of this technology can not only improve the efficiency and accuracy of ocean scientific research but also support the development and utilization of ocean resources and environmental monitoring and protection work. The main purpose of path planning is to find one or more optimal driving routes given a starting point and an ending point. These routes need to meet the requirements of being safe, effective, and as short as possible or conforming to other performance indicators (such as time, energy consumption, smoothness, etc.).

[0003] The A* algorithm has the advantages of both the Dijkstra algorithm and the BFS algorithm. It uses heuristic search, determines the next position according to the value of the cost function, selects the node with the minimum cost as the expansion node, and finally generates a path.

[0004] The ant colony algorithm (Ant Colony Optimization, ACO), also known as the ant algorithm, is a bionic optimization algorithm proposed by Italian scholars Dorigo M et al. in the early 1990s. This algorithm is inspired by the real ant foraging behavior in nature and solves complex optimization problems by simulating the pheromone transmission and path selection mechanisms in the process of ants finding food. It simulates the pheromone released by ants during the foraging process and uses its accumulation and volatilization to guide ants to find the best or optimal solution.

[0005] In the ant colony algorithm, the problem to be solved is abstracted as a path planning problem in graph theory. Nodes represent candidate solutions in the solution space of the problem, and edges represent the conversion relationships between candidate solutions. Each ant randomly selects a starting point in the graph and selects the next moving direction according to the pheromone concentration and heuristic information on the path. Through multiple iterations, the ant colony gradually converges to the optimal solution or an approximate optimal solution. Summary of the Invention

[0006] In view of the defects existing in the prior art, the present invention provides a multi-task point path planning method for an underwater robot considering the marine environment. Different from the traditional ant colony algorithm, this method can solve the performance problems of the traditional ant colony algorithm, such as slow convergence speed and falling into local optimum. On the basis of integrating the A* algorithm, an adaptive pheromone factor, heuristic function factor, and evaporation factor are used for the traditional ant colony algorithm, which can effectively improve the path planning speed. At the same time, multiple indicators are used as the evaluation function, which is of great significance for the efficient and safe operation of the underwater robot.

[0007] The above object is achieved by the following technical solutions:

[0008] A multi-task point path planning method for an underwater robot considering the marine environment, the method comprising the following steps:

[0009] S1: Construct a map according to the grid method, and construct an ocean current function according to the ocean current situation;

[0010] S2: According to the determined grid, determine the evaluation system for path planning, use the ocean current function in step S1 to complete the energy consumption construction, and then complete the smoothness construction, and linearly weight to form the corresponding fitness function;

[0011] S3, according to the current environment model, set the starting point and multiple target points, and use the A* algorithm to find the shortest distance between any target points;

[0012] S4: After calculating the distance between any two target points, transfer the distance matrix to the improved ant colony algorithm, initialize the relevant parameters of the ant algorithm, and set the number of ants and the maximum number of iterations;

[0013] S5: After initializing the parameters, further improve the ant colony algorithm using an adaptive pheromone factor, heuristic function factor, and evaporation factor;

[0014] S6: On the basis of completing S5, use the roulette method to move the ants until the target point is reached. After all ants complete one iteration process, update the pheromone according to the fitness function determined in step S2;

[0015] S7, repeat the iteration process of S6 until all ants reach the maximum number of iterations, output the optimal path planning result, and obtain the corresponding path length, energy consumption value, and smoothness value.

[0016] Preferably, the specific method for map modeling in step S1 is:

[0017] Using a grid mesh, mark the meshes with obstacles as obstacle meshes, representing areas where the robot cannot enter; mark the obstacle-free meshes as blank meshes to indicate the range where the robot can enter. The motion model of the underwater robot in the ocean is as follows:

[0018]

[0019] Among them, θ is the angle of the underwater robot's navigation speed on the θ axis, and u and v represent the components of the ocean current along the x-axis and y-axis directions respectively.

[0020] Preferably, the specific method of step S2 includes:

[0021] Adopt a multi-objective function S composed of three indicators: the path length L p and the energy consumption E p and the path smoothness M p , adopt a fitness function of linear summation, and improve the pheromone update rule: p

[0022] S p = k1L p + k2E p + k3M p

[0023] In the formula, k1 is the weight factor of the path length, k2 is the weight factor of the energy consumption, and k3 is the weight factor of the path smoothness.

[0024]

[0025] In the formula, n is the total number of nodes, p i is the i-th node on the path p; p i+1 is the (i + 1)-th node on the path p; L(p i , p i+1 ) is the Euclidean distance from node i to node i + 1 on the path p;

[0026] Let the energy consumption of the underwater robot moving from node i to j be denoted as J(i, j), and the simplified model of J(i, j) is

[0027]

[0028] Among them, T represents the time required for movement, V represents the speed of the underwater robot. Assuming that the underwater robot moves at a constant speed and the flow rate and direction of the water current are constant, then:

[0029]

[0030] Take E p ​= ∑J(i,j),(s…,i,j,…,g), where s is the starting node and g is the target node, representing the total cumulative navigation energy consumption of the underwater robot when affected by ocean currents during movement, d ij represents the distance between two nodes;

[0031] At the same time, the path smoothness parameter is used to ensure that the underwater robot can avoid collisions as much as possible during operation. During the movement of the underwater robot, the angle difference is θ, and the smoothness degree is expressed by referring to the angle difference of the underwater robot during movement.

[0032]

[0033] Among them, (x i , y i ) is the current position of the underwater robot, (x i-1 , y i-1 ) is the position of the underwater robot at the previous moment, and (x i+1 , y i+1 ) is the position at the next moment.

[0034] Preferably, the specific method described in step S3 includes:

[0035] f(n) = g(n) + h(n)

[0036] Among them, f(n) is the evaluation fitness function of the node, g(n) is the path distance from the current point to the starting point, and h(n) represents the estimated path distance between the target point and the current point. Its expression is:

[0037]

[0038] Among them, (x0, y0) is the coordinate value of the current node, (x s , y s ) is the coordinate value of the starting position, and (x g , y g ) is the coordinate value of the end position.

[0039] Preferably, in step S4, the number of ants and the maximum number of iterations are set as follows: The maximum number of iterations is set to 100 and the number of ants is 50 according to multiple repeated experiments.

[0040] Preferably, the specific method of step S5 includes:

[0041] An adaptive pheromone factor is adopted. When the number of iterations is relatively small, the pheromone factor is relatively small, and as the number of iterations increases, the pheromone factor gradually increases, so that the movement direction of the ants is mainly determined by the generated pheromone:

[0042]

[0043] In the formula, α0 is the minimum value of the pheromone factor, and α max is the maximum value of the pheromone factor, λ1 is the pheromone factor change parameter, and t is the number of iterations;

[0044] An adaptive heuristic function factor β is adopted, and β decreases as the number of iterations increases, satisfying that the value of β is relatively small in the later stage of iteration:

[0045]

[0046] In the formula, β0 is the minimum value of the heuristic function factor, and β min is the minimum value of the heuristic function factor, λ2 is the heuristic function factor change parameter, and t is the number of iterations;

[0047] Adjust the pheromone evaporation factor. By using the method of improving the pheromone evaporation factor, the pheromone evaporation factor ρ decreases as the number of iterations increases. In the dynamic adjustment, ρ min is the minimum value of the pheromone evaporation factor,

[0048]

[0049] Use the global optimal method to update the pheromone where S best is the multi-objective index of the optimal path in this iteration,

[0050]

[0051] Among them, Q represents the coefficient of pheromone change, which is a constant, and λ3 represents the parameter for adjusting the pheromone factor.

[0052] Preferably, the specific method of step S6 includes using the roulette method to determine the moving position of the next target point of the ant

[0053]

[0054] where τ ij (t) is the pheromone content between the positions of the t-th iteration on the grid map, that is, the pheromone content from node i to j, and η ij (t) is the heuristic function, α is the pheromone factor, β is the heuristic function factor, and allow k is the node allowed to pass next time, and d ij represents the Euclidean distance from position i to j;

[0055] When all the ants have traversed all the target points, they will update the remaining pheromone of each path. These pheromones are all on the path (i, j). The pheromone on the path (i, j) is adjusted and updated at the (t + 1)-th iteration:

[0056]

[0057]

[0058] Among them, ρ is the pheromone evaporation factor, and the range of ρ is 0 < ρ < 1; Δτ ij (t) represents the increment of pheromone on the path (i,j) in this cycle, and Δτ ij k (t) is the amount of information that the k-th ant stays on this path at the current moment. Q represents the coefficient of pheromone change, and L k represents the total distance of the route traveled by the k-th ant in the current cycle

[0059] Beneficial effects:

[0060] The present invention provides a method for path planning of an underwater robot with multiple task points considering the ocean environment. Different from the traditional ant colony algorithm, this method can solve the performance problems of the traditional ant colony algorithm such as slow convergence speed and falling into local optimum. On the basis of integrating the A* algorithm, an adaptive pheromone factor, heuristic function factor, and evaporation factor are used for the traditional ant colony algorithm, which can effectively improve the path planning speed. At the same time, multiple indicators are used as evaluation functions, which is of great significance for the efficient and safe operation of the underwater robot. Brief description of the drawings

[0061] The drawings constituting a part of this application are used to provide a further understanding of this application, making other features, purposes, and advantages of this application more obvious. The schematic embodiments and descriptions of the drawings of this application are used to explain this application and do not constitute an improper limitation of this application.

[0062] In addition, throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the elements and elements are not necessarily drawn to scale.

[0063] In the drawings;

[0064] Figure 1 : Flowchart of the method for path planning of an underwater robot with multiple task points considering the ocean environment proposed by the present invention.

[0065] Figure 2 : Schematic diagram of the grid method modeling proposed by the present invention.

[0066] Figure 3 : Schematic diagram of the ocean current modeling method proposed by the present invention.

[0067] Figure 4 : Schematic diagram of the method for calculating the path smoothness proposed by the present invention.

[0068] Figure 5 : Flowchart of the improved ant colony algorithm proposed by the present invention.

[0069] Figure 6 : Flowchart of path planning using the fused A* algorithm and the improved ant colony algorithm proposed by the present invention.

[0070] Figure 7: Path planning results corresponding to the optimal path length, minimum energy consumption, and optimal smoothness on a 30x30 map using the improved ant colony algorithm proposed by the present invention.

[0071] Figure 8: Path planning results corresponding to the optimal path length, minimum energy consumption, and optimal smoothness on a 50x50 map using the improved ant colony algorithm proposed by the present invention.

[0072] Figure 9 : Comparison between the results of the improved ant colony algorithm proposed by the present invention and the path distance convergence graph of the traditional ant colony algorithm. Detailed implementation manner

[0073] The following is a detailed description of the embodiments of the present invention: These embodiments are implemented on the premise of the technical solution of the present invention, and detailed implementation manners and specific operation processes are given. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention.

[0074] It is worth mentioning that unless otherwise specified, the technical terms used in the present invention application should have the ordinary meaning understood by those skilled in the art to which the present invention belongs.

[0075] The present invention provides an underwater robot multi-task point path planning method considering the ocean environment. Different from the traditional ant colony algorithm, this method can solve the performance problems of the traditional ant colony algorithm such as slow convergence speed and falling into local optimum. On the basis of the algorithm integrating the A* algorithm, the expression of the heuristic function of the traditional ant colony is changed, and adaptive pheromone factor, heuristic function factor, and evaporation factor are used, which can effectively improve the path planning speed. At the same time, multiple indicators are used as evaluation functions, which is of great significance for the efficient and safe operation of underwater robots.

[0076] An underwater robot multi-task point path planning method considering the ocean environment according to the present invention, the method comprising the following steps:

[0077] S1: Construct a map according to the grid method, and at the same time construct an ocean current function according to the ocean current situation;

[0078] S2: Based on the determined grid, determine the evaluation system for path planning. Use the ocean current function in step S1 to complete the construction of energy consumption, and then perform the construction of smoothness. Linearly weight to form the corresponding fitness function;

[0079] S3. According to the current environmental model, set the starting point and multiple target points, and use the A* algorithm to obtain the shortest distance between any target points;

[0080] S4: After completing the distance calculation between any two target points, transfer the distance matrix to the improved ant colony algorithm, initialize the relevant parameters of the ant algorithm, and set the number of ants and the maximum number of iterations;

[0081] S5: After initializing the parameters, further improve the ant colony algorithm using adaptive pheromone factor, heuristic function factor, and evaporation factor;

[0082] S6: On the basis of completing S5, use the roulette wheel method to move the ants until the target point is reached. After all ants complete one iteration process, update the pheromone according to the fitness function determined in step S2;

[0083] S7. Repeat the iteration process of S6 until all ants reach the maximum number of iterations, output the optimal path planning result, and obtain the corresponding path length, energy consumption value, and smoothness value.

[0084] Preferably, the specific method of the map modeling in step S1 is:

[0085] Use an accurate grid grid, mark the grids with obstacles as obstacle grids (black part), representing the areas that the robot cannot enter; mark the grids without obstacles as blank grids (blank area) to indicate the range that the robot can enter. The modeling in this article is as Figure 2 shown:

[0086] In addition to being affected by the navigation distance, the energy consumption of the underwater robot during actual navigation is also affected by the ocean current. Therefore, in the overall planning, the role of the water flow cannot be ignored. Since this article only focuses on underwater robots for large-scale navigation, the motion model under ocean current disturbance is simplified, and on this basis, an energy consumption coefficient is introduced to establish a disturbance model of the ocean current on the system. Among them, the motion model of the underwater robot in the ocean is as Figure 3 shown:

[0087]

[0088] Among them, θ is the angle of the underwater robot's navigation speed on the θ axis, and u and v represent the components of the ocean current along the x-axis and y-axis directions respectively.

[0089] Preferably, the specific method of step S2 includes:

[0090] The path length L p and energy consumption E p and path smoothness M p The multi-objective function S composed of these three indicators p ,This multi-objective function aims to plan a comprehensive path with the shortest path length, the least energy consumption and the best smoothness. A linear summation fitness function is used, and the pheromone update rule is improved.

[0091] S p =k1L p +k2E p +k3M p

[0092] Where k1 is the weight factor of path length, k2 is the weight factor of energy consumption, and k3 is the weight factor of path smoothness.

[0093]

[0094] In the formula, n is the total number of nodes, p is i is the i-th node on path p; p i+1 is the i+1th node on path p; L(p i ,p i+1 ) is the Euclidean distance from node i to node i+1 on path p.

[0095] Assume that the energy consumption of the underwater robot moving from node i to node j is recorded as J(i,j), and the simplified model of J(i,j) is:

[0096]

[0097] Where T represents the time required for movement, V represents the speed of the underwater robot, and assuming that the underwater robot moves at a constant speed and the flow rate and direction of the water are constant, then:

[0098]

[0099] Take E p =∑j(i,j),(s…,i,j,…,g), s is the starting node, g is the target node, and it represents the total navigation energy consumption of the underwater robot when it is affected by the ocean current during movement. ij Indicates the distance between two nodes.

[0100] At the same time, the path smoothness parameter is used to ensure that the underwater robot can avoid collisions as much as possible during operation. Figure 4 The θ represents the degree of smoothness by citing the angle difference of the underwater robot during the motion process.

[0101]

[0102] Among them, (x i , y i ) is the current position of the underwater robot, (x i-1 , y i-1 ) is the position of the underwater robot at the previous moment, (x i+1 , y i+1 ) is the position at the next moment, and M p is used to represent the degree of corner smoothness of the underwater robot during movement.

[0103] Preferably, the specific method of step S3 includes: The specific method described in step S3 is as follows:

[0104] f(n) = g(n) + h(n)

[0105] Among them, f(n) is the evaluation fitness function of the node, g(n) is the path distance from the current point to the starting point, and h(n) represents the estimated path distance between the target point and the current point. Its expression is:

[0106]

[0107] Among them, (x0, y0) is the coordinate value of the current node, (x s , y s ) is the coordinate value of the starting position, and (x g , y g ) is the coordinate value of the end position.

[0108] The traditional A* algorithm expands in four neighborhoods, namely, up, down, left, and right, with a minimum turning angle of 90 degrees; and an eight-neighborhood expansion, which adds four directions of upper right, lower right, upper left, and lower left, and the minimum turning angle is 45 degrees. The underwater robot moves slowly underwater and has weak maneuverability. Therefore, it is not suitable for large-angle rotational movements. The eight-neighbor expansion algorithm is used to implement the movement of the underwater robot.

[0109] Preferably, the specific method of step S4 includes: Set the maximum number of iterations to 100 and the number of ants to 50 according to multiple repeated experiments.

[0110] Preferably, the specific method of step S5 includes:

[0111] Adopt an adaptive pheromone factor. When the number of iterations is relatively small, make the pheromone factor relatively small, and as the number of iterations increases, the pheromone factor gradually increases, so that the movement direction of the ants is mainly determined by the generated pheromone.

[0112]

[0113] In the formula, α0 is the minimum value of the pheromone factor, α max is the maximum value of the pheromone factor, λ1 is the pheromone factor change parameter, and t is the number of iterations.

[0114] An adaptive heuristic function factor β is adopted, and β decreases as the number of iterations increases, satisfying that the value of β is relatively small in the later stage of iteration:

[0115]

[0116] In the formula, β0 is the minimum value of the heuristic function factor, β min is the minimum value of the heuristic function factor, λ2 is the heuristic function factor change parameter, and t is the number of iterations.

[0117] Adjust the evaporation factor. In the initial stage, a larger value is adopted to increase the search range of ants and improve the possibility of solution; in the later stage, a smaller evaporation factor is used to improve the convergence speed. The present invention uses a method of improving the pheromone evaporation factor, so that the pheromone evaporation factor ρ decreases as the number of iterations increases, improving the convergence speed while ensuring the path diversity. In the dynamic adjustment formula, ρ min is the minimum value of the evaporation factor.

[0118]

[0119] Use the global optimal method to update the pheromone, where S best is the multi-objective index of the optimal path in this iteration.

[0120]

[0121] Preferably, the step S6 specifically includes: the specific method of the step S5 includes using the roulette method to determine the moving position of the next target point of the ant

[0122]

[0123] Among them, τ ij (t) is the pheromone content between the positions of the t-th iteration on the grid map, that is, the pheromone content between moving from position i to j, η ij (t) is the heuristic function, α is the pheromone factor, β is the heuristic function factor, allow k is the node allowed to pass next time, d ij represents the Euclidean distance between moving from position i to j.

[0124] After all ants have traversed all target points, they will update the remaining pheromones on each path, which are all on path (i, j). Usually, the pheromone adjustment and update on path (i, j) at the (t + 1)-th iteration are as follows:

[0125]

[0126] Among them, ρ is the pheromone evaporation factor, and the range of ρ is 0 < ρ < 1; represents the increment of pheromone on path (i, j) in this cycle, the amount of information when the k-th ant stays on this path at the current moment, Q represents the coefficient of pheromone change, and L k represents the total distance of the route traveled by the k-th ant in the current cycle.

[0127] Preferably, step S7 is to obtain the path result of this planning and the required path distance, energy consumption, and smoothness after the process is completed.

[0128] This method is a new path planning result of an underwater robot based on the ant colony algorithm under the traditional ant colony algorithm, which can be directly applied in the marine environment and considered in the path planning of underwater robots under various constraints.

[0129] Compiled using the Matlab language, it uses the Windows 11 operating system with an i9-14900, a main frequency of 2.4 GHz, and a running memory of 16 GB. To make the experimental conclusions more representative and accurate, experiments were repeated in two environments with fewer obstacles of 30x30 and more obstacles of 50x50. Among them, the parameter settings are α0 = 1, α max is 6, β0 is 6, β min is 1, ρ min is 0.2, λ1 = 0.5, = 0.5, λ3 = 1.

[0130] In the 30x30 environmental model, according to the actual marine environment, the ocean current factor is added. Assuming that the AUV travels at a constant speed V = 1 m / s, the ocean speed component along the x-axis u is 0.3 m / s, and the component along the y-axis v is 0.4 m / s. The paths obtained when k1, k2, and k3 take different proportional values are shown in Figures 7(a) to (c):

[0131] Table 1 Path results under different weight coefficients in 30x30

[0132]

[0133] As can be seen from Figure 7, when the multi-objective function weight focuses on the path length, the path order at this time is S-4-2-1-3-5-7-6-8-S. When the multi-objective function weight focuses on the energy consumption, the path order at this time is S-2-4-8-1-3-5-7-6-S. When the multi-objective function weight focuses on the path smoothness, the path order at this time is S-2-1-3-5-6-7-8-4-S. Display the results of Figure 7 in Table 1, and it can be seen that the results meet the preset performance. Different optimal sequences are obtained according to different optimization conditions.

[0134] In the 50x50 environment model, the environment is more complex than that of 30x30 at this time, and the parameter settings are the same as those of 30x30. The results shown in Figure 8 are obtained as follows: among them, (a) is the sequence curve corresponding to the least path distance, and (b) is the sequence curve corresponding to the least energy consumption. (c) is the sequence curve corresponding to the path smoothness.

[0135] To verify the effectiveness of the algorithm, the number of target points is increased to 9 in the dense environment. As can be seen from the results of Figure 8, when the multi-objective function weight focuses on the path length, the path order at this time is S-7-8-9-6-5-3-2-1-4-S. When the multi-objective function weight focuses on the energy consumption, the path order at this time is S-6-5-3-2-1-9-8-7-4-S. When the multi-objective function weight focuses on the path smoothness, the path order at this time is S-6-5-4-1-2-3-9-8-7-S. Display the results of Figure 8 in Table 2, and it can be seen that the results meet the preset performance. Different optimal sequences are obtained according to different optimization conditions.

[0136] Table 2 Path results under different weight coefficients in 50x50

[0137]

[0138] At the same time, by setting the comparison with the traditional ant colony algorithm in terms of performance, the set target points are the same as those of 30x30. The parameter settings of the traditional ant colony algorithm are α = 1.5, β = 2, and ρ = 0.9, but the weight coefficient is only k1 = 1. The comparison graph of the convergence curves of the algorithm in this paper and the traditional algorithm is as Figure 9 shown: The improved result not only speeds up the convergence rate but also avoids falling into the local optimum.

[0139] In summary, combining the attached drawings and the above table, the algorithm proposed in this embodiment can significantly speed up the convergence rate and at the same time obtain a better distance compared with the traditional ant colony algorithm.

[0140] This embodiment provides an underwater robot path planning method based on an improved ant colony algorithm considering the marine environment. Different from the traditional ant colony algorithm, this method can solve the performance problems of the traditional ant colony algorithm, such as slow convergence speed and falling into local optimum. On the basis of the algorithm, an adaptive pheromone factor, heuristic function factor, and evaporation factor are used, which can effectively improve the path planning speed. At the same time, multiple indicators are used as the evaluation function, which is of great significance for the efficient and safe operation of underwater robots.

[0141] Finally, it should be noted that the above embodiments are only a detailed clarification of the technical solution and not a limitation thereof; although the technical solution of the present invention has been described with reference to the specific embodiments, those of ordinary skill in the art should understand that they can still modify the solutions of the embodiments or make equivalent substitutions for some of them; such modifications or substitutions do not deviate from the scope of the technical solution proposed by the present invention in essence and should all be included in the scope of the claims and the specification of the present invention.

Claims

1. A multi-task point path planning method for an underwater robot considering the marine environment, characterized in that: The method comprises the following steps: S1: Map construction is carried out according to the grid method, and the ocean current function is constructed according to the ocean current conditions; S2: Based on the determined grid, the evaluation system of the path planning is determined. The energy consumption is constructed using the ocean current function of step S1, followed by the smoothness construction, and the linear weighting is used to form the corresponding fitness function. The specific method includes: The path length L p and energy consumption E p and path smoothness M p The multi-objective function S composed of these three indicators p , using a linear summation fitness function and improving the pheromone update rule: S p =k1L p +k2E p +k3M p In the formula, k1 is the weight factor of path length, k2 is the weight factor of energy consumption, and k3 is the weight factor of path smoothness; In the formula, n is the total number of nodes, p is i is the i-th node on path p; p i+1 is the i+1th node on path p; L(p i ,p i+1 ) is the Euclidean distance from node i to node i+1 on path p; Assume that the energy consumption of the underwater robot moving from node i to node j is recorded as J(i,j), and the simplified model of J(i,j) is: Where T represents the time required for movement, V represents the speed of the underwater robot, assuming that the underwater robot moves at a uniform speed, θ represents the angle between the underwater robot's navigation speed and the θ axis, u and v represent the components of the ocean current along the x-axis and y-axis respectively; and the velocity and direction of the water flow are constant, then: Take E p =∑J(i,j),(s…,i,j,…,g), s is the starting node, g is the target node, and it represents the total navigation energy consumption of the underwater robot when it is affected by the ocean current during movement, d ij Represents the distance between two nodes; At the same time, the path smoothness parameter is quoted to ensure that the underwater robot can avoid collision as much as possible during operation. During the movement of the underwater robot, the angle difference is θ. The degree of smoothness is expressed by quoting the angle difference of the underwater robot during the movement. Among them, (x i ,y i ) is the current position of the underwater robot, (x i-1 ,y i-1 ) is the position of the underwater robot at the last moment, (x i+1 ,y i+1 ) is the position at the next moment; S3, according to the current environment model, set the starting point and multiple target points, and use the A* algorithm to find the shortest distance between any target points; S4: After completing the distance calculation between any two target points, the distance matrix is ​​passed to the improved ant colony algorithm, the relevant parameters of the ant algorithm are initialized, and the number of ants and the maximum number of iterations are set; S5: After initializing the parameters, the adaptive pheromone factor, heuristic function factor, and volatile factor are further used to improve the ant colony algorithm; the specific method is: Adopting adaptive pheromone factor, when the number of iterations is relatively small, the pheromone factor is relatively small, and as the number of iterations increases, the pheromone factor gradually increases, so that the movement direction of the ants is mainly determined by the generated pheromones: In the formula, α0 is the minimum value of the pheromone factor, α max is the maximum value of the pheromone factor, λ1 is the pheromone factor variation parameter, and t is the number of iterations; Adopting the adaptive heuristic function factor β, β is reduced as the number of iterations increases, so that the β value is relatively small in the later stage of iteration: In the formula, β0 is the minimum value of the heuristic function factor, β min is the minimum value of the heuristic function factor, λ2 is the change parameter of the heuristic function factor, and t is the number of iterations; Adjust the pheromone volatility factor. Use the method of improving the pheromone volatility factor to make the pheromone volatility factor ρ decrease as the number of iterations increases. In the dynamic adjustment, ρ min is the minimum value of pheromone volatility factor, Use the global optimal method to update pheromones Where S best is the multi-objective indicator of the optimal path in this iteration, Among them, Q represents the coefficient of pheromone change, which is a constant, and λ3 represents the parameter for adjusting the pheromone factor; S6: On the basis of completing S5, the ants are moved by using the roulette method until they reach the target point. After all ants have completed one iteration, the pheromone is updated according to the fitness function determined in step S2; S7, repeat the iterative process of S6 until all ants reach the maximum number of iterations, output the optimal path planning result, and obtain the corresponding path length, energy consumption value and smoothness value.

2. The method for multi-task point path planning of an underwater robot considering the marine environment according to claim 1, characterized in that: The specific method of constructing the map in step S1 is: Using a square grid, grids with obstacles are marked as obstacle grids, representing areas that the robot cannot enter; grids without obstacles are marked as blank grids to indicate the range that the robot can enter. The motion model of the underwater robot in the ocean is as follows: Among them, θ is the angle of the underwater robot's navigation speed on the θ axis, and u and v represent the components of the ocean current along the x-axis and y-axis respectively.

3. The method for multi-task point path planning of an underwater robot considering the marine environment according to claim 1, characterized in that: The specific method of step S3 includes: f(n)=g(n)+h(n) Among them, f(n) is the evaluation fitness function of the node, g(n) is the path distance from the current point to the starting point, and h(n) represents the estimated path distance between the target point and the current point. Its expression is: Among them, (x0, y0) is the coordinate value of the current node, (x s ,y s )The coordinate value of the starting position, (x g ,y g ) is the coordinate value of the end point.

4. The method for multi-task point path planning of an underwater robot considering the marine environment according to claim 1, characterized in that: The step S4 sets the number of ants and the maximum number of iterations: the maximum number of iterations is set to 100 and the number of ants is set to 50 based on repeated experiments.

5. The method for multi-task point path planning of an underwater robot considering the marine environment according to claim 1, characterized in that: The specific method of step S6 includes using the roulette method to determine the moving position of the ant's next target point Among them, τ ij (t) is the pheromone content between the t-iteration positions on the grid map, that is, the pheromone content between moving from node i to j, η ij (t) is the heuristic function, α is the pheromone factor, β is the heuristic function factor, allow k is the node that is allowed to pass next time, d ij Represents the Euclidean distance from position i to j; When all ants have traversed all target points, they will update the remaining pheromones on each road. These pheromones are all on the path (i, j). The pheromones on the path (i, j) are adjusted and updated at the t+1 iteration: Among them, ρ is the pheromone volatility factor, and the range of ρ is 0<ρ<1; Δτ ij (t) represents the increment of pheromone in path (i, j) in this cycle, Δτ ij k (t) is the amount of information that the kth ant stays on the path at the current moment, Q represents the coefficient of pheromone change, and L k Represents the total distance of the route traveled by the kth ant in the current cycle.

Images