Soccer Robot Simulation System Based On The Behavior Of Fuzzy Control

Robot soccer system has now become the important experiment platform for the technical research of artificial intelligence. A decision-making system is used as its central part to extract knowledge and to control the robot to accomplish the task cooperatively. In a dynamic environment where opponent exists, mutual cooperation and group behavior are needed by the robot to achieve a specific goal. Meanwhile, because of the dynamic of the environment and the unpredictability of the opponent, higher demands are given to the real-time decision-making system and control of the robot.This paper takes its research target on the five-versus-five game in the FIRA robot soccer match. This paper, based on the structure of the simulation system and the movement modal of the robot, analyzes the theories of the multi-agent system and the characteristics of the movement control of the mobile robot. In the decision-making system, the cooperation of the agents and the movement control of the mobile robot are deeply analyzed and researched in order to overcome the dynamic and unpredictability of the environment.(1) From the perspective of behavior, a double-layer decision-making model is putted forward with the upper layer being the specialist cooperation agent based on deduction and the under layer being the movement control agent based on actions. The upper-layer purpose design and the under-layer action command of the robots are completely divided to enhance the flexibility of the robot and the consistence of the decision-making system.(2) A fuzzy control algorithm based on behavior is used to make the robot to a specific point. And a linear-velocity fuzzy controller and an angel fuzzy controller are designed to combine the fuzzy control's robustness with the "feel—action" behavior based on physiology. This presents a new idea for the control of the mobile robot in an unknown environment.(3) By comparing the results of simulation experiment between fuzzy control algorithm and the traditional PID algorithm, we can seethat the fuzzy control algorithm can greatly improve smoothness and levelness of robot movement.