The Research And Application Of The Algorithm On Mobile Robot Path Planning

Mobile robots have the ability to resolve difficult tasks in complicated environments,they have been used in many fields such as explosive ordnance disposal, space exploring,healthcare and so on. Planning out a complete path is a basic requirement to make mobilerobots been used in many fields successfully. Therefore, more and more scholars began to doresearch on path planning.This paper researches on the algorithms of mobile robot path planning, the algorithm is ablended algorithm that combines artificial potential field method with rapidly randomexploring trees. The main content includes:1. This paper analyses the basic principles, the advantages and disadvantages of somecommonly used path planning algorithm. Especially, focused on analyzing artificial potentialfield method, the traditional method has the problem of can’t get the goal and fall into localstable region easily. To solve this problem, this paper amends repulsion potential function andbuilds middle coordinates. The improved algorithm is only suitable for the condition that thearound environment is simple and the obstacles arrange regularly.2. When the environment is complicated, the improved artificial potential field methodis still falling into local stable region easily. To solve this problem, this paper proposes ablended algorithm which combines rapidly random exploring trees with artificial potentialfield method. RRT algorithm samples randomly can get artificial potential field method rid oflocal stable region.3. RRT algorithm samples randomly and searches the whole environment uniformly,not taken the solution cost into account during the search process. Then, the algorithm is poorof application, the final path isn’t optimal. To solve this problem, using the anytime algorithmto improve RRT algorithm. Initially, the approach generates a set of useful rapidly exploringrandom trees rapidly. Then, removing useless sample points before each iteration. This workensures that the path is better than the previous one.4. When a lot of narrow passages exist in complex environment, the hybrid algorithmcouldn’t sample in narrow passages, the result path isn’t optimal. Bridge test algorithm canget the samples easily in the narrow passages, this thesis improves the hybrid algorithm byusing bridge test algorithm. The improved hybrid algorithm applies in RoboCup.