Collaborative Research: Dynamics of Running on Variable Inclines

开发能够在平坦、倾斜、不均匀和可变形表面上有效运行的有腿机器人的能力对于广泛的民用（如搜索和救援）和军事（如侦察、地雷探测）应用至关重要。对于传统的控制策略来说，这项任务已经被证明是特别困难的，然而许多动物，不像它们的人造对手，能够非常轻松地超越复杂的环境，如丛林、洞穴、沙漠、建筑物和成堆的瓦砾。尽管它们的体型很小，但昆虫几乎可以在任何环境中发挥作用，它们可以根据情况的需要爬、爬或跑。尽管最近在能够在不平坦的地面上移动的仿生机器人的开发方面取得了进展，其他机器人也可以（缓慢地）爬上陡峭的表面，但目前还没有一种有腿的机器人能够在垂直和水平的组合中动态地工作。最近发现，快速攀爬的蟑螂和壁虎利用它们的腿主动将身体拉向脚，而不是像跑步时那样将身体推开，这启发了垂直奔跑的新动态模型和第一个动态攀爬机器人的构建。这种攀爬模型和机器人的显著特点是在快速攀爬时有意利用大的侧向拉力。由于在实验中观察到的这些拉力运动似乎没有提供明显的能量优势，我们假设这种类型的侧到侧攀爬主要是出于稳定性考虑。本项目综合研究昆虫生物力学、动力学建模和机器人合成，以确定横向振荡在不同倾斜度的奔跑和攀爬中的重要性和正确利用。这项研究旨在回答这些问题，不仅为动物生物力学提供见解，而且还为开发下一代动态有腿机器人提供指导原则。特别是，我们的目标是了解昆虫的神经肌肉控制策略与变化环境（即斜坡和基质）中的功能表现之间的联系，开发降阶模型以确定横向运动模式如何有助于提高稳定性和运动表现。并开发了一个新的机器人测试平台，以经验测试仿生运动模型的预测，并为如何明确利用横向动力学来提高有腿机器人在不同地形和基质上的机动性提供见解。

The ability to develop legged robots that can effectively run over flat, inclined, uneven and deformable surfaces is critical to a wide range of civilian (e.g., search and rescue) and military (e.g., reconnaissance, mine detection) applications. This task has proven particularly problematic to traditional control strategies, yet many animals, unlike their synthetic counterparts, are able to transcend complex environments such as jungles, caves, deserts, buildings, and piles of rubble with remarkable ease. Despite their miniscule size, insects can function in almost any environment by means of their ability to climb, crawl, or run as the situation demands. Although recent progress has been made in the development of bio-inspired robots capable of locomotion over uneven ground, and others that can (slowly) climb sheer surfaces, no legged machines currently exist that can dynamically operate in a combination of vertical and horizontal regimes. The recent discovery that rapidly climbing cockroaches and geckos utilize their legs to actively pull the body toward the feet, rather than pushing the body away as in running, has inspired a new dynamic model for vertical running and the construction of the first dynamic climbing robot. The salient feature of this climbing model and robot is the intentional utilization of large lateral pulling forces when rapidly climbing. Since these pulling motions observed experimentally do not appear to provide an obvious energetic advantage, we hypothesize that this type of side to side climbing is driven primarily by stability considerations. This project pursues an integrated study of insect biomechanics, dynamic modeling, and robotic synthesis to determine the importance and proper utilization of lateral oscillations in running and climbing over various degrees of incline. The study seeks to answer these questions not only to provide insight into animal biomechanics, but also to produce guiding principles that can be utilized to develop the next generation of dynamic legged robots. In particular we aim to understand the connection between neuromuscular control strategies in insects and functional performance in changing environments (i.e. slope and substrate), develop reduced order models to determine how lateral motion pattern contribute to improved stability and locomotion performance, and develop a novel robotic test platform to empirically test the predictions of the bio-inspired locomotion models and provide insight into how lateral dynamics can be explicitly utilized to improve the mobility of legged robots over varying terrain and substrates.