Sampling-based hierarchical motion planning for a reconfigurable wheel-on-leg planetary analogue exploration rover
Sampling-based hierarchical motion planning for a reconfigurable wheel-on-leg planetary analogue exploration rover
复制标题
DOI：
10.1002/rob.21894
复制
发表时间：
2019-10-22
影响因子：
8.3
通讯作者：
Sukkarieh, Salah
中科院分区：
文献类型：
--
作者：
Reid, William;Fitch, Robert;Sukkarieh, Salah
可重构的移动的行星漫游车是多功能平台，可以通过改变其物理几何形状来安全地穿越杂乱的环境。这些自适应机器人的规划路径是具有挑战性的，由于他们的许多自由度，并需要考虑潜在的连续平台重构沿着的长度的path. We提出了一种新的分层结构的渐近最优（AO）采样为基础的规划者，并特别适用于它的国家的最先进的快速行进树（FMT*）AO规划。我们的算法假设一个完整的配置空间分解成多个子空间，并开始通过快速找到一组路径通过这样的子空间。这组解决方案用于生成有偏采样分布，然后探索该分布以在全配置空间中找到解决方案。这种技术提供了一种新的方式，将先验知识的子空间，有效地偏置搜索现有的AO采样为基础的规划。重要的是，概率完备性和渐近最优性得到了保持。仿真实验结果表明，基准的算法对国家的最先进的采样为基础的规划，没有层次的变化。附加的实验结果进行了一个物理轮腿平台演示应用行星漫游车的流动性，并展示如何约束，如致动器故障和传感器指向可以很容易地纳入规划问题。在最小化的能量目标，结合了近似的机械工作所需的平台运动与所需的重新配置，规划器产生直观的行为，机器人动态调整其足迹，改变其高度，并爬上障碍物使用腿运动。这些结果说明了规划者在利用平台的机械能力来在各种物理几何配置和轮式/腿式运动模式之间流体地过渡而不需要预定义配置方面的一般性。
Reconfigurable mobile planetary rovers are versatile platforms that may safely traverse cluttered environments by morphing their physical geometry. Planning paths for these adaptive robots is challenging due to their many degrees of freedom, and the need to consider potentially continuous platform reconfiguration along the length of the path. We propose a novel hierarchical structure for asymptotically optimal (AO) sampling-based planners and specifically apply it to the state-of-the-art Fast Marching Tree (FMT*) AO planner. Our algorithm assumes a decomposition of the full configuration space into multiple subspaces, and begins by rapidly finding a set of paths through one such subspace. This set of solutions is used to generate a biased sampling distribution, which is then explored to find a solution in the full configuration space. This technique provides a novel way to incorporate prior knowledge of subspaces to efficiently bias search within existing AO sampling-based planners. Importantly, probabilistic completeness and asymptotic optimality are preserved. Experimental results in simulation are provided that benchmark the algorithm against state-of-the-art sampling-based planners without the hierarchical variation. Additional experimental results performed with a physical wheel-on-leg platform demonstrate application to planetary rover mobility and showcase how constraints such as actuator failures and sensor pointing may be easily incorporated into the planning problem. In minimizing an energy objective that combines an approximation of the mechanical work required for platform locomotion with that required for reconfiguration, the planner produces intuitive behaviors where the robot dynamically adjusts its footprint, varies its height, and clambers over obstacles using legged locomotion. These results illustrate the generality of the planner in exploiting the platform's mechanical ability to fluidly transition between various physical geometric configurations, and wheeled/legged locomotion modes, without the need for predefined configurations.