基于新人工势场函数的机器人动态避障规划

Dynamic obstacle-avoiding path plan for robots based on a new artificial potential field function

人工势场法是进行机器人路径规划时常用的方法,但若用圆锥曲线函数作为引力场数学模型时,在目标点会产生抖动问题.本文在分析抖动产生原因的基础上,增加一个指数项到引力场函数中,从而消除了奇异值点,避免了抖动现象.然后将一敏感度参数引入斥力场函数,以便灵活控制运动过程中机器人与障碍物距离的大小.通过对敏感度的调节,还可以克服传统势场法中目标点在斥力作用范围内时,机器人无法到达目标点的缺陷.最后给出新势场法的仿真.

The artificial potential field is in the common use to plan the robot path.But if conic curve function is the model of attractive field, there is chattering phenomenon at target point. Based on the analysis of chattering cause, a new potential field function is presented by adding an exponential factor to the conic curve attractive function. It can eliminate singularity chattering in planning. Meanwhile, in order to achieve flexible control of distance between robot and obstacles, a parameter called sensitivity is introduced in the repulsive function. By adjusting the magnitude of the sensitivity, the deficiency that robot cannot reach its target inside the range of repulsion action in traditional potential field function can be avoided. Simulation results are given.