机器人路径规划的遗传算法求解策略分析

Analysis of Solving Strategy of Genetic Algorithm for Robot Route Planning

针对遗传算法求解有障碍物的机器人路径规划问题,不同的遗传选择策略,求解性能各有差异.为此,构建平面内含有不同障碍物的四种测试场景,障碍物分别选取矩形、圆形及其混合图形,在相同实验参数下,分别采用轮盘赌策略、君主策略、锦标赛策略三种选择策略进行求解.经过重复实验,统计所求最短路径长度的均值和标准差,结果表明锦标赛策略比其他策略求得的最短路径长度均值和标准差都更小.因此,锦标赛策略在这类问题求解中具有更好的求解性能.

For solving the robot route planning problem with obstacles by the genetic algorithm,different genetic selection strategies have different performances.Therefore,four test scenarios with different obstacles in the plane are constructed.Obstacles are selected from rectangles,circles and their mixed graphics.Under the same experimental parameters,the three selection strategies of roulette wheel selection,emperor selection and tournament selection are used to solve.After repeated experiments,the mean and standard deviation of the shortest path lengths are counted.The results show that the mean and standard deviation of the shortest path lengths obtained by the tournament selection are smaller than those of other selections.Therefore,the tournament selection has better performance in solving such problems.