Magnetic Cogging Parallel-elastic Actuators for Energy-efficient Robotic Legs

该奖项为下一代机器人探索了一种新颖的使能技术，该技术灵活、高效且能够安全地与人类互动：磁性齿槽平行弹性执行器（MC-PEA）。 MC-PEA 特别有希望用于机器人腿，包括假肢、外骨骼以及人形机器人和四足机器人等自主机器人的腿。 MC-PEA 包括与无源磁齿槽扭矩元件 (CTE) 并联的电动机。 MC-PEA 的决定性特征是它具有由强无源磁弹簧创建的一组离散的、可设计控制的稳定平衡。电机不需要能量来简单地保持静态负载。相反，需要爆发能量才能使 CTE 达到邻近的平衡。 MC-PEA 将有效地实现仿生运动原语，包括多种共振模式，这些模式将导致步行和跑步期间所需的高效振荡运动。在目前 SEA 是首选解决方案的应用中，MC-PEA 将成为普遍存在的串联弹性执行器 (SEA) 的节能替代品。此外，该奖项还将支持研究生的培训和指导、本科生参与研究以及教学课程的改进。该资助的目的是表征磁齿槽平行弹性致动器（MC-PEA）的自然和受迫动力学特征，并确定如何有效且高效地控制它以实现机器人运动。这项工作将增进对优化设计的 MC-PEA 跨潜在参数变化的性能规格的理解。 MC-PEA 的优势将通过在动力踝假肢和腿式机器人中的应用来探索。该方法包括对仿生动态运动基元的研究，包括阻抗控制、点对点子运动和振荡，使用带有简单点状脚的两连杆机器人腿，并使用 MC-PEA 为臀部和膝盖提供动力。该研究还包括开发定制的能量再生电路，该电路可以快速打开和关闭，以便在动态运动期间回收能量。该项目的成果将是各种步长的磁齿槽扭矩元件 (CTE) 的最佳参数设计，并将分发给机器人设计人员。该项目得到跨部门机器人基础研究项目的支持，该项目由工程理事会 (ENG) 和计算机与信息科学与工程 (CISE) 共同管理和资助。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

This award explores a novel enabling technology for next-generation robots that are agile, efficient, and that safely interact with humans: the magnetic cogging parallel-elastic actuator (MC-PEA). The MC-PEA is particularly promising for use in robotic legs, including prostheses, exoskeletons, and the legs of autonomous robots such as humanoids and quadrupeds. The MC-PEA comprises an electric motor connected in parallel with a passive magnetic cogging-torque element (CTE). The defining characteristic of the MC-PEA is that it has a discrete, controllable-by-design set of stable equilibria created by strong passive magnetic springs. No energy is required from the motor to simply hold a static load. Rather, bursts of energy are required to cause the CTE to cog over to a neighboring equilibrium. The MC-PEA will efficiently implement bioinspired motion primitives, including multiple resonance modes that will lead to efficient oscillatory movements required during walking and running. The MC-PEA will represent an energy-efficient alternative to the ubiquitous series-elastic actuator (SEA) in applications in which the SEA is currently the go-to solution. In addition, this award will support training and mentorship of graduate students, inclusion of undergraduate students in research, and improvements to teaching curriculum.The objective of this grant is to characterize the natural and forced dynamics of the magnetic cogging parallel-elastic actuator (MC-PEA) and determine how to control it effectively and efficiently for robotic motions. The work will improve understanding of the performance specifications of an optimally designed MC-PEA across potential parameter variations. The benefits of the MC-PEA will be explored through application in powered ankle prosthesis and legged robots. The approach includes investigation of bioinspired dynamic motion primitives—including impedance control, point-to-point sub-movements, and oscillations—using a two-link robotic leg with a simple point-like foot and with MC-PEAs powering the hip and the knee. The research also includes development of custom energy regeneration circuits that can be rapidly switched on and off to enable energy to be reclaimed during dynamic movements. The project outcomes will be the optimal parametric designs of the magnetic cogging-torque element (CTE) for a wide variety of step sizes, to be disseminated to robot designers.This project is supported by the cross-directorate Foundational Research in Robotics program, jointly managed and funded by the Directorates for Engineering (ENG) and Computer and Information Science and Engineering (CISE).This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.