Efficient collective shape shifting and locomotion of massively-modular robotic structures
Efficient collective shape shifting and locomotion of massively-modular robotic structures
复制标题
大规模模块化机器人结构的高效集体变形和运动
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
P. Hołobut
中科院分区：
文献类型：
--
作者：
J. Lengiewicz;P. Hołobut
提出了一种基于立方晶格的大型集成模块化机器人的有效变形和运动规划方法。这些模块分为两组：固定模块，构建一个刚性的多孔框架；移动模块，通过框架流动。从结构中流出的可移动模块附着在框架上，推进其边界。相反，在边界的其他部分移动模块的不足是通过分解框架来纠正的。在结构内部，通过最大流搜索算法的一个特殊的分布式版本，安排适当的模块流，使模块按期望的方向运输。该方法在重新配置过程中使用了大量模块，比普通的表面流方法更有效。此外，建议的解释为具有移动边界的多孔介质中的流动似乎特别适合进一步开发更先进的全局重构场景。首先对该方法的理论有效性进行了评估，然后通过一系列仿真验证了该方法的部分有效性。这种方法也可以应用于更广泛的模块化机器人，不一定是基于立方格的。
We propose a methodology of planning effective shape shifting and locomotion of large-ensemble modular robots based on a cubic lattice. The modules are divided into two groups: fixed ones, that build a rigid porous frame, and mobile ones, that flow through the frame. Mobile modules which flow out of the structure attach to the frame, advancing its boundary. Conversely, a deficiency of mobile modules in other parts of the boundary is corrected by decomposition of the frame. Inside the structure, appropriate module flow is arranged to transport the modules in a desired direction, which is planned by a special distributed version of a maximum flow search algorithm. The method engages a volume of modules during reconfiguration, which is more efficient than common surface-flow approaches. Also, the proposed interpretation as a flow in porous media with moving boundaries seems particularly suitable for further development of more advanced global reconfiguration scenarios. The theoretical efficiency of the method is assessed, and then partially verified by a series of simulations. The method can be possibly also applied to a wider class of modular robots, not necessarily cubic-lattice-based.