极坐标系下机器人路径规划算法

Robot path planning algorithm in polar coordinate

针对机器人避障问题,提出一种基于遗传算法的极坐标系下机器人路径规划算法,将传统直角坐标系下采用遗传算法的路径规划改为极坐标系下的路径规划方法;给出极坐标系下的运动模型,引入有障碍物模式和无障碍物模式.其中有障碍物模式采用遗传算法,简化适应度函数,加入重启策略,根据单纯型思想提出新的交叉算子;改进后的算法分为直接向目标点行进和采用遗传算法寻找最短轨迹2种情况.仿真实验表明:设起始点距离为100～500cm时,该算法与传统直角坐标系下采用遗传算法进行路径规划比较平均运行时间节省75.9%,平均规划路径长度缩短20.8%,适应度函数仅有1个约束条件,验证算法的有效性、实时性,降低决策的盲目性.

Focusing on the problem of robot avoiding obstacles, a path planning method of robot in Polar coordinate based on genetic algorithm is proposed. A movement model in Polar coordinate is given and obstacles mode and no obstacles mode are introduced here. The traditional path planning which uses genetic algorithm in rectangular coordinate is improved to path planning method in Polar coordinate. Using the genetic algorithm in obstacles model, fitness function is simplified, restart strategy is introduced, according to the simplex method, new crossover operator is proposed. The improved algorithm is divided into two cases, one travels to the goal directly, the other uses the genetic algorithm to seek the shortest path. Compared with the traditional path planning which uses genetic algorithm in rectangular coordinate, it can save time by 70%, cut the planning length by 20%. The simulation results demonstrate the effectiveness and real-time and reduces the blindness of the proposed algorithm.