复杂动态环境下自主机器人路径规划研究

Automatic Robot Path Planning Under Complicit Dynamic Environment

为了提高复杂动态环境下的机器人路径规划性能,文章提出了全局路径规划和局部路径规划相融合的混合路径规划方法。分析了A*算法原理,提出了加权A*算法,使用权值调节启发信息在评价函数中的作用;改进了人工势场法,解决了传统方法目标不可达和局部极值问题;将加权A*算法的全局路径规划和改进人工势场法的局部路径规划相融合,以全局规划路径的拐点作为局部路径规划的子目标点,提出了混合路径规划方法。经仿真验证,对于固定静态障碍物、临时堆放障碍物、动态障碍物三种情况,混合规划方法都能得到平滑的无碰最优路径。

To improve property of robot path planning under complicit dynamic environment,path planning method combined global planning and local planning method is proposed. A^＊algorithm principle is analyzed,and weighted A^＊algorithm is put forward. The weight can adjust magnitude of heuristic information in evaluation function. Artificial potential field is improved,goal unreachable and local extremum are solved. Global path planning by weighted A^＊algorithm and local path planning by improved artificial potential field are combined,and inflection point of global path is regarded as sub goal of local path,so that combined path planning method is present. By simulation,under condition of fixed static obstacle,temporary placed obstacle,and dynamic obstacle,smooth optimal without touching path can be planned by combined path planning method.