Research And Application On Path Planning Based On Dynamic Artifical Potential Field

In recent years, path planning problem is a hot issue of robot technology area and an important branch of robot study in artifical intelligence. This thesis is based on robot combat games, which mainly include Microsoft wheel micro robot simulation game platform(MSRS) and the overall vision robot for water polo match platform. Against the competition system for robot path planning problem unique to high complexity, real-time, reflecting the demand for prompt and multi-agent characteristics, a path planning method based on dynamic artificial potential field is proposed in the paper. And it is mainly to solve the problem that the destination can not be reached and the local minimum trap problem when using traditional artificial potential field method for path planning, and finally that path planning method is written to the ultimate decision-making in the real game, which achieved good results. The main work of this paper are:Firstly, this paper provides an introduction to the background of robot system, mobile robot path planning research and the research practical significance. The paper provides a introduction to Microsoft wheel micro soccer robot simulation game platform and the overall vision robot for water polo match platform including some rules which are used in the experiment.Secondly, the thesis makes an analysis of the traditional artificial potential field model and the main issue, in response to these issues advance the corresponding countermeasures, which include against the robot can not reach the destination problem to inprove the repulsive consitute of potential field model and use virtual water-flo to solve the local minimum trap problem.And both are done simulation contrast experiment in the MATLAB platform.Lastly, the improved artifical potential field called dynamic artifical potential field is used in path planning.Thirdly,the paper provides a particular introduction to decision-making constitutes of MSRS. The path planning method with the dynamic artifical potential field is used in the decision-making of MSRS, analyze actual game data compared with other planning method, such as Genetic Algorithm, Ant Colony Algorithm and General Geometric Algorithm. examine that path planning method in the overall vision robot for water polo match platform, then analyze and compare the effect of different experimental platform.Actual race results show that a small amount of calculation, simplify calculation method and taking into account the performance of the robot movement in trajectory planning are the features of the path planning method with the dynamic artifical potential field. So it has better result in avoiding obstacle in robot combat game, and can be well positioned to meet the complexity and real-time requirements of the game.