CN-121977560-A - Dynamic multi-target path planning method for autonomous underwater robot cooperation task

CN121977560ACN 121977560 ACN121977560 ACN 121977560ACN-121977560-A

Abstract

The invention discloses a dynamic multi-target path planning method for an autonomous underwater robot cooperative task. Aiming at the conflict between communication and task optimization in an underwater cooperative task and the influence of the highly dynamic change of a communication blind area on a path planning solution space, a pareto solution set optimization strategy and a single solution strategy mechanism based on historical path adaptability relation prediction are provided. Firstly, constructing a communication target optimization function based on a potential field method, guiding an AUV to avoid or rapidly separate from a communication blind area, and ensuring stable and reliable communication. And secondly, predicting the distribution of the optimal solution based on the fitness relation by analyzing the evolution trend of the historical solution set, optimizing the formation mode of the initial solution, and improving the solving efficiency of the genetic algorithm. And finally, selecting a unique solution from the pareto solution set based on a packet loss rate decision mechanism, overcoming the inaccuracy of the existing decision index, and ensuring that the optimized result has more engineering feasibility and practical application value.

Inventors

LIU KAIZHOU
WANG YINHUAN
GENG LINGBO
ZHANG SHAOZE

Assignees

中国科学院沈阳自动化研究所

Dates

Publication Date: 20260505
Application Date: 20260121

Claims (9)

1. A dynamic multi-target path planning method for an autonomous underwater robot cooperative task is characterized by comprising the following steps: 1) Discretizing task space based on a grid method; 2) Constructing a communication optimization objective function based on a potential field method; 3) According to the communication quality optimization objective function, constructing a multi-objective optimization function; 4) Taking the multi-objective optimization function as a fitness function, predicting the distribution of the optimal solution, constructing an initial solution set, and obtaining a pareto solution set through a genetic algorithm; 5) And selecting a unique solution from the pareto solution set based on a packet loss rate decision mechanism as a final path.
2. The method for planning the dynamic multi-objective path for the cooperative task of the autonomous underwater robot according to claim 1, wherein the task space is discretized based on a grid method, and the method comprises the following steps: firstly, acquiring position information of a task target point and a communication relay node, wherein the position information is respectively used as a path planning terminal point and a communication optimization target node; Secondly, taking the current position of the AUV as a path planning starting point; Finally, determining the size and density of the grid based on a grid method in combination with a path planning starting point, a path planning end point and a space range of a communication optimization target node; The task area is Divide it into A plurality of grid cells, each grid marked as , wherein, , , Representative grid Coordinates of the center; Representing the length and width of the task area respectively, Representing the length and width of the grid cell, respectively.
3. The method for planning a dynamic multi-objective path for cooperative tasks of an autonomous underwater robot according to claim 1, wherein the constructing a communication optimization objective function based on a potential field method comprises the following steps: when constructing a communication potential field, taking an autonomous underwater robot currently bearing a communication relay task as a reference AUV, and recording the current grid as The movement direction is The communication radius is According to the local coordinate system with the position of the reference AUV as the origin In the direction of the AUV movement direction Two candidate communication traction points G1 and G2 are symmetrically arranged at the position, the distance between the candidate traction points and the current position of the AUV needing to plan a path is calculated respectively, and the traction point with smaller distance is selected as a preferable communication traction point and used for actively guiding the AUV to move to a region with better communication quality and separate from a communication blind region; the gravitational potential field generated by the traction points is calculated by a potential field method, each AUV constructs a traction point potential field of the AUV on the local layer, and the gravitational field function is based on the grid where the AUV is currently in communication relay And a target grid Calculated according to the following formula: Representing the potential field of the pull-in point, The grid is represented by a grid of lines, Represents the coefficient of the gravitational field, Representing a two-dimensional euclidean norm; in the discretized task space, the communication relay AUV is taken as the center of a circle and the communication radius is taken as the communication radius A sector area with a preset included angle of 30 DEG symmetrical to the tail axis is defined as a communication blind area, and the distance between the sector area and the relay AUV is smaller than that between the sector area and the relay AUV To give an obstacle potential value 10 times the maximum value of the traction point potential field Forming a repulsive force potential field of a communication blind area, wherein the AUV is larger than or equal to the AUV To give a communication interruption penalty value of 10 times the maximum value of the traction point potential field Establishing a communication interruption potential field for characterizing an area in which a communication link cannot be established; by superposing the traction point potential field, the blind area repulsive force potential field and the communication interruption potential field, a combined communication potential field is formed ; ; In the discretized task space, the path of motion of the AUV is expressed as Then the communication optimizes the objective function Expressed as: Wherein, the Representation of The corresponding grid, n, represents the total number of path points.
4. The method for planning a dynamic multi-objective path for cooperative tasks of an autonomous underwater robot according to claim 1, wherein the optimizing objective function according to the communication quality constructs a multi-objective optimizing function, comprising the steps of: constructing task efficiency optimization objective function by calculating total length of connecting line from starting point to target point The calculation formula is as follows: Wherein, the The coordinates of the starting point are represented, Indicating the point of the i-th path, Representing the distance between two points; combining task efficiency optimization and communication quality optimization targets to construct a multi-target optimization function The mathematical expression is as follows: Wherein, the For the number of path points, Representing a combined communication potential field.
5. The method for planning the dynamic multi-objective path for the cooperative task of the autonomous underwater robot according to claim 1, wherein the steps of taking the multi-objective optimization function as a fitness function, predicting the distribution of the optimal solution, constructing an initial solution set, and obtaining the pareto solution set through a genetic algorithm include the following steps: (1) Screening and cutting out the gene segments of the current environment from the historical solution set through knowledge extraction: Historical solution set as , wherein, Represents a historical Pareto-optimal solution, Representing a decision variable; Will be Bringing into multi-objective optimization functions Then, a binary group is obtained Wherein A sequence of path points is represented and, Representing the objective function value under the path; cutting path points of the historical solution set in the current environment by taking the point closest to the end point as a reference to obtain a gene fragment as the current solution; (2) Multi-objective optimization function The evolution of the solution is predicted based on the adaptability variation trend of the historical solution as a function of the adaptability, and the POFs in the historical solution are classified into two layers by adopting a non-dominant ordering method And ; (3) Combining knowledge hierarchy strategies with inflection point prediction methods to construct new initial populations: Will pass through And Inflection point of prediction And from And The method comprises the steps of organizing solutions of the population into a new initial population according to the fitness function value sequence, randomly supplementing the population if the size of the population does not reach a threshold value, selecting an optimal subset through priority ranking if the size of the population exceeds the threshold value, and then optimizing the new population by adopting NSGA-II and obtaining a current pareto solution set.
6. The method for planning the dynamic multi-objective path for the cooperative task of the autonomous underwater robot according to claim 1, wherein the packet loss rate-based decision mechanism selects a unique solution from the pareto solution set as a final path, specifically comprising the following steps: When packet loss rate When the threshold value is exceeded, selecting a solution with optimal communication objective function from the pareto solution set: Wherein, the The unique solution for the screening is represented, Representing the communication optimization objective function, Representing the sum of the distances of all the path point links.
7. A dynamic multi-objective path planning system for collaborative tasks of autonomous underwater robots, comprising: The discretization module is used for discretizing the task space based on a grid method; The optimization function construction module is used for constructing a communication optimization objective function based on a potential field method; the pareto solution set acquisition module is used for taking the multi-objective optimization function as an fitness function, predicting the distribution of the optimal solution, constructing an initial solution set, and obtaining the pareto solution set through a genetic algorithm; And the path selection module is used for selecting a unique solution from the pareto solution set based on a packet loss rate decision mechanism and taking the unique solution as a final path.
8. The dynamic multi-target path planning device for the autonomous underwater robot cooperation task is characterized by comprising a memory and a processor, wherein the memory is used for storing a computer program, and the processor is used for realizing the dynamic multi-target path planning method for the autonomous underwater robot cooperation task according to any one of claims 1-6 when the computer program is executed.
9. A computer readable storage medium, wherein a computer program is stored on the storage medium, and when the computer program is executed by a processor, a dynamic multi-objective path planning method for collaborative tasks of an autonomous underwater robot is implemented according to any of claims 1-6.

Description

Dynamic multi-target path planning method for autonomous underwater robot cooperation task Technical Field The invention belongs to the field of path planning of autonomous underwater robots, and particularly relates to a dynamic multi-target path planning method for cooperative tasks of an autonomous underwater robot. Background In the underwater collaborative task, the autonomous underwater robot (Autonomous Underwater Vehicle, AUV) not only needs to adjust the spatial distribution of the group according to the topology optimization result to optimize the group communication quality, but also needs to consider the completion efficiency of the collaborative search task. Complex constraints of the marine environment, uncertainty of target motion and limited communication capability also need to be considered in the process, so that the problem of autonomous underwater robot path planning in an underwater collaborative search task is a typical dynamic multi-target optimization problem. However, the existing research does not fully consider the influence of communication blind areas when constructing an objective function, so that the stability of communication is difficult to ensure in practical application of a path planning scheme. In addition, due to the rapid change of the communication blind area, the distribution of the infeasible area can be dynamically adjusted, so that the solution space of the multi-objective optimization problem is highly dynamic. The existing research is difficult to meet the requirement of quick and accurate solution, and the adaptability and the robustness of the path planning method in a complex task environment are insufficient. Disclosure of Invention The invention aims to provide a dynamic multi-objective path planning method for AUV cooperative tasks. Aiming at the conflict between communication and task optimization in an underwater cooperative task and the influence of the highly dynamic change of a communication blind area on a path planning solution space, a pareto solution set optimization strategy and a single solution strategy mechanism based on historical path adaptability relation prediction are provided. Firstly, constructing a communication target optimization function based on a potential field method, guiding an AUV to avoid or rapidly separate from a communication blind area, and ensuring stable and reliable communication. And secondly, predicting the distribution of the optimal solution based on the fitness relation by analyzing the evolution trend of the historical solution set, optimizing the formation mode of the initial solution, and improving the solving efficiency of the genetic algorithm. And finally, selecting a unique solution from the pareto solution set based on a packet loss rate decision mechanism, overcoming the inaccuracy of the existing decision index, and ensuring that the optimized result has more engineering feasibility and practical application value. The technical scheme adopted by the invention for realizing the purpose is that the dynamic multi-target path planning method for the cooperative task of the autonomous underwater robot comprises the following steps: 1) Discretizing task space based on a grid method; 2) Constructing a communication optimization objective function based on a potential field method; 3) According to the communication quality optimization objective function, constructing a multi-objective optimization function; 4) Taking the multi-objective optimization function as a fitness function, predicting the distribution of the optimal solution, constructing an initial solution set, and obtaining a pareto solution set through a genetic algorithm; 5) And selecting a unique solution from the pareto solution set based on a packet loss rate decision mechanism as a final path. The grid-based method for discretizing the task space comprises the following steps: firstly, acquiring position information of a task target point and a communication relay node, wherein the position information is respectively used as a path planning terminal point and a communication optimization target node; Secondly, taking the current position of the AUV as a path planning starting point; Finally, determining the size and density of the grid based on a grid method in combination with a path planning starting point, a path planning end point and a space range of a communication optimization target node; The task area is Divide it intoA plurality of grid cells, each grid marked as, wherein,,,Representative gridCoordinates of the center; Representing the length and width of the task area respectively, Representing the length and width of the grid cell, respectively. The communication optimization objective function is constructed based on a potential field method, and the communication optimization objective function comprises the following steps: when constructing a communication potential field, taking an autonomous underwater robot currently beari