一种基于粒子群算法的轮式机器人路径规划

A Method of Path Planning for Wheeled Robot Based on Particle Swarm Optimization

针对轮式机器人具有非完整性约束的情况,为了优化机器人路径,提出了利用粒子群算法处理轮式机器人在滚动窗口内的路径优化问题。根据机器人当前滚动窗口内的障碍物信息进行避障和趋向目标运动,同时在充分考虑轮式机器人运动特性的基础上构建了一个符合其非完整性约束和避障要求的适应度函数,从而能较好地满足实时避障的要求。为了尽量减少可行区域由于障碍物建模造成的损失,提出了采用改进的栅格法进行建模,该方法较全面地考虑了障碍物的外形尺寸。仿真实验证明上述方法的正确性和有效性。

There is constraint of non -holonomy in wheeled robot, so the method of path optimization explored in this paper is to utilize particle swarm algorithm to settle the problem in rolling window. By means of this method, the wheeled robot can do obstacle -avoidance and object -oriented movement only according to the obstacle information of the robot in the present rolling window, at the same time a fitness function for the robot which satisfies its non - ho- lonomy and obstacle -avoidance requirements is constructed on the basis of sufficiently considering the motion fea- tures of the robot, which thereby makes the robot do better to satisfy the requirements of the timely obstacle - avoid- ance. The paper proposes improved grids method in order to increase area movement of the robot, according to the shape and size of the obstacle. Simulate experiments have demonstrated its accuracy and availability.