Improved Path-Finding Algorithm for Robot Soccers

In robot soccer game mobile robot moves in complex, unpredictable, unknown dynamic environment. In such environment the design of efficient path finding algorithm that will find the path in a short time becomes important problem. The widely used path finding methods are potential field method, vector field histogram and A* algorithms. In real time operations finding the optimal path in a reasonable short time is not possible with the mentioned methods. The RRT (Rapidly Exploring Random Tree) algorithm finds a path in short time. But some times the length of path may be very long. This paper is devoted to the design of improved path finding algorithm that will find the near optimal path in a short time. The improved iterative RRT with path smoothing algorithm is developed and implemented in RoboCup Small Size robots. Through simulation it was shown that this algorithm can efficiently find desirable and near optimal solutions in short time. 