基于改进型遗传算法的足球机器人路径研究

目前针对足球机器人路径规划,主要采用栅格法和链接图法,但栅格法无法达到精确的规划路径,而连接图法主要针对具有复杂多边形的障碍物,这两种方法不能满足足球机器人实时性的要求.为此采用简化编码方式和格雷码,建立了以路径最短、避障为优化目标的遗传算法适应度函数,采用轮盘赌选择,单点交叉,基本位变异等方法,完成了遗传操作.仿真结果表明:在建立的约束关系下,改进型遗传算法在路径最短方面比人工势场法有所改进,表现出较好的优化效果.

Optimized Path-planning in Soccer Robot System Based on an Improved Genetic Algorithm

At present, there are two main path planning methods for the soccer robot. One is grid method, the other is MAKLINK graph.But the accurate planned path can not be achieved by grid method, and the MAKLINK graph is used to solve the complicated polygonal roadblock problems, so neither of them can meet the real-time requirement of soccer robot control system. In order to solve the problem, this paper presents an adaptive genetic algorithm function by applying a simplified code and Gray codes in control algorithm, by which the soccer robot can move along the shortest path and avoid obstacle. And by the roulette wheel selection, one-point crossover and simple mutation, the genetic operation has been completed. The simulation results show that the improved genetic algorithm has better characteristics in path planning than the artificial potential method.