Active Dynamic Continuum Tails for Maneuvering and Stabilizing Legged Robots

这项研究的结果将为制造机器人尾巴提供机会，这种尾巴能够增强有腿机器人在非结构化危险环境中敏捷而稳健地穿越地形的能力。该研究将影响从搜索和救援到侦察到探索危险和危险环境的应用，在这些环境中，机器人必须在非结构化地形中有效移动。与轮式或履带式系统相比，腿式机器人可以更好地穿越这些非结构化环境，因为腿使用了离散的地面接触点。这个项目的主要影响来自于尾巴对有腿机器人的稳定性和可操作性的改善。这将导致更快的搜索、救援或危险环境的探索，使这些有腿的机器人能够增强并超越人类的能力。在教育影响方面，本研究将：(1)为工程研究生和本科生提供支持，(2)发展为大学水平课程的案例研究，(3)用于K-12学生的示范。少数民族和代表性不足的群体将成为外联和参与的对象。该奖项的研究目标是研究仿生连续体尾巴如何动态稳定和操纵高速移动的有腿移动机器人。野外机器人在如何应对非结构化危险环境中不可预见的动态干扰以及如何在不改变步态和减速的情况下改变方向方面受到限制。有了连续的尾巴，稳定和操纵机器人所需的力和力矩可以通过动态尾巴运动来应用，而不是使用腿、躯干或手臂。本研究将研究将所需的力和力矩映射到连续尾控制输入的算法，以稳定和操纵有腿机器人，并推导动态模型来预测尾巴的动态力和力矩，从而实现任务规划。可交付成果包括智能机器人尾巴的动态建模、任务规划算法和设计范式、尾巴机构测试平台的实验验证和腿部运动仿真、研究成果的传播、工程学生教育、学生和科学教师的工程研究经验以及工程新课程的开发。

Results of this research will provide an opportunity to create robotic tails that are capable of augmenting a legged robot's ability for agile and robust terrain traversal in unstructured hazardous environments. The research will impact applications from search and rescue to reconnaissance to exploring hazardous and dangerous environments, in which robots must move effectively in unstructured terrains. Legged robots can better traverse these unstructured environments compared to wheeled or tracked systems due to the legs' use of discrete ground contact points. This project's primary impact stems from the tail's improvements to a legged robot's stability and maneuverability. This will lead to faster search and rescue or exploration of dangerous environments, allowing these legged robots to augment and exceed human capabilities. In terms of educational impact, this research will: (1) provide support for engineering graduate and undergraduate students, (2) be developed into case studies for university-level courses, and (3) be used in demonstrations for K-12 students. Minorities and underrepresented groups will be targeted for outreach and participation.The research objective of this award is to study how biomimetic continuum tails can dynamically stabilize and maneuver legged mobile robots moving at high speeds. Field robots are limited in how they may respond to unforeseen dynamic disturbances in unstructured hazardous environments and how they may change their direction without modifying gait and slowing down. With a continuum tail, the required forces and moments to stabilize and maneuver the robot can be applied by dynamic tail motions instead of using the legs, torso or arms. This research will study algorithms to map the required forces and moments into continuum tail control inputs to stabilize and maneuver a legged robot, and derive dynamic models to predict the tail's dynamic forces and moments to enable task planning. Deliverables include dynamic modeling, task planning algorithms and design paradigms of intelligent robotic tails, an experimental validation with a test platform of the tail mechanism and simulations of legged locomotion, dissemination of research results, engineering student education, engineering research experiences for students and science teachers, and new engineering curriculum development.