新型势场模型及在实时运动规划中应用  

Novel potential field model and its application in real-time motion planning

作　　者：王孙安[1] 袁明新[1] 庄健[1] 陈乃建[1] 

机构地区：[1]西安交通大学机械工程学院,陕西西安710049

出　　处：《计算机工程与设计》2009年第8期1944-1946,1949,共4页Computer Engineering and Design

摘　　要：针对复杂环境中移动机器人的实时运动规划问题,在传统人工势场法基础上,以机器人与动态障碍物的相对位置和相对速度为变量,构建了一种新型的位置和速度势场模型,提出了一种新的自适应势场法。通过变参数约束扰动法解决了势场法中的局部极小问题。实验结果表明,该方法能使机器人选择最佳避障策略实现主动避障,快速地摆脱动态障碍物,规划出光滑路径,从而提高避障效率,较好地解决了动态未知环境中的路径规划问题。To solve the real-time motion planning for robot in complicated environments, a novel potential field model of position and velocity is built, and a new adaptive potential field method is proposed with the parameters of relative position and velocity among robot and dynamic obstacles. The local minima is solved by changing algorithm parameter value and restricting disturbance. The experimental results show that the robot can avoid dynamic obstacles actively, break away dynamic obstacles quickly and plan smooth path. The proposed method can solve the problem of path planning well in unknown dynamic environments.

关 键 词：势场法 实时 运动规划 最佳避障 局部极小 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]