Path planning and control of soccer robot based on genetic algorithm

摘要

Studying the optimization of the soccer robot path planning is significant because of the ever-changing situation of the arena. The system itself has non-linearity. The environment also has the characteristics of time varying, which requires the robot to cooperate with each other. Combining with the characteristics of a soccer robot system, this paper proposes a genetic algorithm for the path planning algorithm of a soccer robot, which enables the robot to find a relatively short path from the point to the target. In this paper, three common path-planning methods are introduced, and the advantages and disadvantages are compared. Then, a path planning method based onS-adaptive genetic algorithm is proposed. The main innovation point is to change the crossover probability and mutation probability in genetic operation. Therefore, that it can change the probability value autonomously with the change of genetic generation and individual in the group. This method can better save the effective individual, make the algorithm converges faster, and get the optimal individual under the same conditions. The path solved byS-adaptive algorithm realizes obstacle avoidance behavior in a short time, and the path is better than the traditional genetic algorithm.