隘路环境下排爆机械臂Multi-RRT路径规划算法

Multi-RRT Path Planner for EOD Manipulators in Narrow Passage

摘要:

排爆机械臂通常工作在类似隘路的环境中,传统的基于采样的路径规划方法在隘路环境下计算效率较低；提出一种采样增强的Multi-RRT路径规划方法,该方法分为两个部分:第一部分为区域随机桥采样法,该方法以采样特征决定局部C空间(configuration space)随机桥测试的参数并采集有限隘路样本；第二部分为Multi-RRT算法,以起始点、目标点和随机桥采样所得点为树根,多树多方向协调生长规划路径；仿真实验结果表明该方法提高了快速扩展随机树算法的计算效率,有效地解决了隘路环境下排爆机械臂的路径规划问题,适用于多维高自由度机器人系统的路径规划。

Abstract:

The explosive ordnance disposal (EOD) manipulators are usually working in narrow passage circumstances. It is inefficient for classic method based on sampling to solve path planning problems in narrow passages. A certain Multi-RRT path planner based on sample enhancing is proposed. The developed technology has two parts:the first part is region based randomized bridge builder (region-RBB) . Sampling attribute are used to determine the parameter of bridge test of local configuration space and finite samples are collected in narrow passages. The second part is Multi-RRT algorithm. The samples collected by region-RBB, the start node and the goal node are initialized as roots of multi-trees which grow harmoniously to plan path of robots. The simulation results show that the proposed method increases the efficiency of rapidly-exploring random tree and solve the path planning problem of EOD manipulators in narrow passages effectively. And it is also appropriate for path planning of high degree of freedom robots in multi-dimension environment.