Research On Optimized Path Planning And Autonomous Obstacle Avoidance Method Of Surface Unmanned Vessel

With the rapid development of the global economy,marine resources have become a new demand for human exploration,and have also become the focus of national defense and diplomacy with other countries.Unmanned Surface Vehicle(USV)related technical research can enable countries to make better use of marine resources for major power diplomacy.USV,therefore,in the aspects such as autonomous safe navigation and autonomous local obstacle avoidance technology has become a problem we need urgently,and these technology research and development and how quickly to put these techniques into industrialization is inevitable problems,these problems needed to resolve are nowadays China in terms of waterway transport business of green and intelligent led need to conquer the research.In this context,a safe and relatively short route was planned for the USV prior to its mission,avoid obstacles in real time,and deal with critical situations.This paper mainly studies two aspects of USV,one is its global path to the workspace before the mission,the other is its safe local obstacle avoidance to unknown obstacles in the navigation process.The main research contents are as follows:Firstly,raster environment modeling is carried out for the workspace before USV sails along the established route planned by the path.The accuracy of environment modeling determines whether the algorithm can plan an established route with high safety and short distance for the whole world.First introduction to environmental modeling in general,and in this paper,using the grid method to rasterize modeling of electronic chart,and introduces in detail the feasible region and division between obstacles,there are obstacles for precise processing method,for the later proposed algorithm for path planning to build the foundation,embody the rigor of article.Secondly,aiming at the shortcomings of traditional ant colony algorithm,a new method to find the starting point before each iteration is proposed,so that the best solution in the iteration can be fully utilized and improved.And combining the idea of artificial potential field method,an optimized ant colony algorithm,Artificial Potential Field-Ant Colony Algorithm is proposed,which strengthens the use of each ant’s search ability,thereby increasing the utilization rate of the best solution and avoiding local optima.The basic performance of the optimized ant colony algorithm and the other three algorithms are compared and simulated experiments to verify the feasibility of the proposed optimization algorithm.Thirdly,the Optimized Particle Colony and Ant Colony Hybrid Algorithm for local autonomous obstacle avoidance is proposed.Because the USV when navigating in the process of performing tasks need to be able to quickly to detect the unknown obstacle avoidance decision-making measures of dynamic obstacles,so the optimization of particle swarm algorithm and the second chapter of the proposed optimization ant colony algorithm is a combination of foster strengths and circumvent weaknesses,use limit algorithm mixing conditions will together two kinds of algorithms of the initial pheromone,Optimized Particle Colony and Ant Colony Hybrid Algorithm is proposed,the diversity of algorithm operation is increased and the test function is used to compare and analyze the algorithm in the following chapters to highlight the superiority of hybrid algorithm.Finally,the dynamic obstacle avoidance problem of USV during navigation is studied through a large number of analytical experiments.Study such issues which need to pass through the USV obstacle avoidance process,maritime obstacle avoidance motion state and the risk of collision avoidance rules,obstacles to rigorous analysis and calculation,and the calculated results for writing USV autonomous obstacle avoidance simulation software research,using the simulation software for this article makes a comprehensive simulation experiments,the proposed hybrid algorithm.The feasibility of the algorithm is verified by the simulation experiments of USV with two tasks and multiple encounter situations.