Fuzzy Logic Path Planner and Motion Controller by Evolutionary Programming for Mobile Robots

moving obstacles (opponent robots) and a moving target (ball). Until now, many navigation methods have been demonstrated for robot soccer. Sim et. al. [2] proposed a four-mode control structure which has different controllers for four divided areas around a ball. Kim et. al. [3,4] proposed the uni-vector field method. Selecting two points, one in front and the other at the back of the ball, the vector angle from these two points to the current position of a robot is calculated. Then the uni-vector field is constructed by multiplying the vector angle by a certain integer. Wong et. al. [5] proposed an optimized fuzzy controller using a genetic algorithm. A fuzzy logic controller (FLC) for mobile robots is designed in hierarchical structure. The designed FLC consists of two levels: the planner level and the motion control level. The planner level generates a path to the destination by avoiding obstacles. The singleton outputs of the planner are obtained by using lines and arc methods. The lower motion control level calculates the robot's wheel velocity so as to follow the path generated by the planner as to the current robot posture. The fuzzy singleton outputs are obtained by heuristics and tuned by evolutionary programming. The applicability of the controller is demonstrated by using robot soccer system. Fuzzy logic is a widely used tool in mobile robotics. The path following and local obstacle avoidance behaviors are the potential fields for the application of fuzzy logic [6-7]. In most of the designs, the desired path is generated by a higher level planner. In this paper, we apply fuzzy logic to this planner level as well as to the motion control level in designing the shooting action controller of soccer robot (Figure 1).