Research On Intelligent Path Planning Algorithm For Underwater Vehicle

Underwater vehicle is an important medium for performing underwater missions.Underwater vehicle path Planning is an important part of underwater vehicle Research.This paper mainly focuses on the global path planning and local path planning of underwater vehicles in complex environments and the route target allocation for underwater vehicle.The purpose is to use intelligent algorithm planning to obtain optimized paths for different scenarios.This paper first studies the global path planning of underwater vehicles in a complex environment.Considering the deficiency of uniform modeling method applied in complex environment,a non-uniform grid model is obtained by dividing search space by quadtree.Based on the non-uniform grid model,the objective function considering path energy consumption,security and executable is designed in this paper.This paper proposes an improved ant colony algorithm based on the non-uniform grid model for the key nodes of the path.The algorithm has the following improvements to the basic ant colony algorithm: using heuristic distances for non-uniform grid models,consider the security of the selected node when selecting a path node,using the Ant Fallback Policy to Solve the Path Deadlock Problem.In this paper,the comparative simulation experiments between uniform grid and non-uniform grid are designed,and the improved ant colony algorithm is compared with several basic ant colony algorithms and A *.Next,the local path planning of underwater vehicle in complex environment is studied.This paper first establishes a rolling path planning framework that considers the non-integral constraints of underwater vehicles.In view of the fact that the basic RRT(rapidly exploring random tree)algorithm does not take into account the distribution of obstacles,the fixed-length extension step and the random expansion direction are adopted,which leads to the problem of low expansion efficiency in complex environment.This paper proposes "node visibility" to describe the degree of difficulty of node expansion,and an improved RRT algorithm based on node visibility is proposed to select node extension scheme.In this paper,several complex environment simulation examples are designed to compare the search performance between the proposed algorithm and the basic RRT algorithm.Finally,the path target assignment problem of multiple underwater vehicles is studied.A mathematical model of path target assignment for multiple underwater vehicles considering the range and number of strikes and the difference between the targets to be hit is established,the distributed cooperative auction algorithm is used to solve the problem in combination with the actual task requirements.Simulation experiments with the number of underwater vehicles and the number of targets are equal and unequal are designed to verify the performance of the algorithm.In summary,this paper mainly studies the problem of global path planning,local underwater vehicle path planning and multi-underwater vehicle path assignment in complex environment.This study can provide some reference for follow-up research.