NRI: Small: Reflex approximation of optimal control for an energy-efficient bipedal walking platform

NRI：小：节能双足步行平台最优控制的反射近似

• 批准号: 1317981
• 负责人: Andy Ruina
• 金额: $ 99.99万
• 依托单位: Cornell University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2013
• 资助国家: 美国
• 起止时间: 2013-09-01 至 2017-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1317981&HistoricalAwards=false
• 关键词: NRI; Small; Reflex; approximation; optimal

The broad research goal of this project is to better understand the mechanics, balance, and energetics of bipedal walking. The first specific goal is to design, build and test a two-legged walking 'platform', a waist-down robot, and its computer controller. One can think of this platform as a legged Segway, for use by others putting on robotic upper bodies with arms and heads. The goal is to match the robustness of existent bipedal robots but with no need for electrical cords or hydraulic tubes and with approximately a 90 percent reduction in energy use. This will be achieved by the development and use of a reflex-based controller that minimizes the needed sensor bandwidth and on-board computation. The second goal is to advance the theories of walking balance and energetics including, especially, basic models for understanding the tradeoffs between robustness, versatility, energy efficiency, sensor accuracy and motor size. Deliverable outcomes include a demonstrable robot with a new 'reflex-based' controller, public access to complete design and control information, basic design formulas for the fundamental design trade-offs, student design and research experience, and inspiring demonstrations for public viewing.If successful, robots based on this platform design could replace wheel and tread based systems used for, and proposed for use for, service robotics. The advantages of legs over wheels include a more anthropomorphic look, for human interaction, and, ultimately with further development, the ability to function on rough terrain. The basic theories of balance and energetics developed in this work will add insight concerning human walking, ultimately helping with the diagnosis and remediation of human disabilities, including aiding the design of assistive and prosthetic devices. The hands-on development work will educate and inspire engineering students working on the project. The demonstrations and demonstration videos will serve as an inspiration to those considering engineering careers.

该项目的主要研究目标是更好地了解双足行走的力学、平衡和能量学。 第一个具体目标是设计、建造和测试一个两足行走的“平台”，一个腰部向下的机器人及其计算机控制器。 人们可以把这个平台想象成一个有腿的赛格威，供其他人使用，用手臂和头部穿上机器人上身。其目标是与现有的双足机器人的鲁棒性相匹配，但不需要电线或液压管，并减少约90%的能源使用。这将通过开发和使用基于反射的控制器来实现，该控制器最大限度地减少了所需的传感器带宽和机载计算。 第二个目标是推进行走平衡和能量学的理论，特别是包括用于理解鲁棒性、多功能性、能源效率、传感器精度和电机尺寸之间的权衡的基本模型。可实现的成果包括一个可演示的机器人与一个新的“反射为基础的”控制器，公众访问完整的设计和控制信息，基本设计权衡的基本设计公式，学生的设计和研究经验，和鼓舞人心的演示公众view.If成功，机器人基于这个平台的设计可以取代车轮和胎面为基础的系统，用于服务机器人，并建议使用。腿相对于轮子的优势包括更拟人化的外观，用于人类互动，并且最终随着进一步的发展，能够在粗糙的地形上发挥作用。在这项工作中开发的平衡和能量学的基本理论将增加对人类行走的见解，最终有助于人类残疾的诊断和补救，包括帮助辅助和假肢设备的设计。动手开发工作将教育和激励工程专业的学生从事该项目。演示和演示视频将作为那些考虑工程职业的灵感。