Research On The AUV Path Planning Method In Complex Environment

With the greatly improvement of the marine status, ocean exploration and development continued to develop in depth. Autonomous navigation capability is the strategic point to AUV, and the path planning technology is the most important part of autonomous navigation system. To a certain extent, the level of path planning indicates the level of the robot's autonomy. This paper summarizes the existing path planning methods and studies the path planning techniques of the autonomous underwater vehicle based on the latest achievements.First, the definition, classification and major planning methods of the path planning for mobile robots were introduced. Path planning refers to in the work environment which has some obstructions, the mobile robot according to a certain performance evaluation criteria, to find a collision‐free path from the beginning to the end. According to the situation of the environmental information, the path planning problem can be divided into two categories: one is global path planning that all the environmental information is known, and the other is local path planning with unknown environmental information. The commonly used path planning methods are grid method, visibility graphs method, artificial potential field method and artificial intelligence methods of genetic algorithm, ant colony algorithm, and so on.Secondly, this paper describes the artificial potential field local path planning algorithm in detail. The inherent defects of the traditional artificial potential field are analyzed and an improved artificial potential field method to these problems is proposed. In the marine environment, taking the impact of ocean currents for autonomous mobile robots into account, we introduce the acting force of the ocean currents to artificial potential field method. The improved method is applied to experience of the underwater robot developed by ourselves and the experimental results demonstrate that the improved algorithm is of timeliness and effectiveness.Then, the hybrid method combined the artificial potential field algorithm with the ant colony algorithm for robot path planning is analyzed in this paper. We introduce the basic principles and specific program of ant colony algorithm firstly, on this basis a hybrid method combined the local artificial potential field path planning algorithm with the ant global colony algorithm path planning for robot is proposed and the simulation experiments are carried out for it. Through analyzing the experimental results for the new algorithm, it proves the feasibility of the method. The new hybrid algorithm avoids the local maxima of artificial potential field method and optimizes the ant colony algorithm so that it can find the optimal path planning.Finally, this paper discusses and summarizes the existing techniques of mobile robot path planning problem and then the development trends are preliminarily discussed of the path planning technology under the unknown and complex environment.