CAREER: Artificial Muscle Based on Dielectric Elastomers for Dexterous and Compliant Prostheses

该教师早期职业发展（CAREER）项目的主要目标是实现具有与天然肌肉相似的机械和动态特性的人工肌肉。特别是，该项目考虑使用一类称为介电弹性体（DE）的材料，这些材料具有与生物肌肉相当的顺应性，弹性和单位面积的力。像生物肌肉一样，这些材料可以自我感知，允许精确控制收缩或伸展，而不需要视觉反馈或其他辅助传感方案。该项目的最终目标是使用基于DE的人工肌肉实现灵巧、轻便和节能的假肢，与当前一代机器人手臂的笨重和低效的电动马达形成对比。该项目结合了生物启发设计，设备制造，动态建模，传感和控制等方面。该项目的成功将有助于为约200万因事故、自然灾害、战争、疾病或衰老而失去手、臂或腿的退伍军人和平民提供负担得起、可靠和舒适的假肢。该项目还将培训下一代劳动力，掌握基于智能材料和结构的动态建模，控制和制造设备的技能。吸引学生进入这一研究领域的活动将提高科学、技术、工程和数学（STEM）学科的入学率。本研究的长期目标是开发轻量级、顺应性和自感知的DE，以模仿机器人辅助应用中生物肌肉的致动和感知。实现这一目标的第一步是获得生物启发的设计，建模，自感知和控制策略的DE执行器在假肢手的应用。本项目的核心目标是：1）建立一种新型的人工肌肉结构，该结构由管状DE人工肌肉和碳纤维人工肌腱组成; 2）建立一个基于低阶物理的模型，该模型描述了DE致动器的非线性弹性和应变相关电阻抗;（3）利用致动器对应变相关电阻抗的自感知，实现了一种灵敏的鲁棒非线性状态观测器;（4）推导出状态边界回避控制策略以保护DE致动器免受损坏;以及（5）构建DE使能的假手以展示具有高效且顺应性致动的灵巧操纵。

This Faculty Early Career Development (CAREER) project has the main goal of realizing artificial muscle with mechanical and dynamic properties similar to natural muscle. In particular, this project considers the use of a class of materials called dielectric elastomers (DEs) that have compliance, resilience, and force per area comparable to biological muscles. Like biological muscles, these materials can be self-sensing, allowing precise control of contraction or extension without needing visual feedback or other auxiliary sensing schemes. The ultimate goal of the project is to achieve dexterous, lightweight, and energy-efficient prostheses using DE-based artificial muscles, in contrast to the heavy and inefficient electric motors of the current generation of robotic arms. The project incorporates aspects of bio-inspired design, device fabrication, and dynamic modeling, sensing, and control. The success of this project will help provide affordable, reliable, and comfortable prostheses to the estimated two million military veterans and civilians who have lost hands, arms, or legs to accidents, natural disasters, wars, diseases, or aging. This project will also train the next-generation workforce with skills in the dynamic modeling, control, and fabrication of devices based on smart materials and structures. Activities to attract students to this area of research will improve enrollment in science, technology, engineering, and mathematics (STEM) disciplines.The long-term goal of this research is to develop lightweight, compliant, and self-sensing DEs to emulate the actuation and sensing of biological muscles for robotic assistive applications. A first step towards this goal is to obtain bio-inspired design, modeling, self-sensing, and control strategies for DE actuators in a prosthetic hand application. This project will emphasize the following core goals: 1) create a novel artificial muscle structure consisting of a tubular DE artificial muscle attached to carbon fiber artificial tendons; 2) derive a low-order physics-based model capturing the nonlinear elasticity and strain-dependent electrical impedance of the DE actuator; (3) implement a sensitive and robust nonlinear state observer using actuator self-sensing of the strain-dependent electrical impedance; (4) derive a state-boundary avoidance control strategy to protect the DE actuator from damage; and (5) build a DE-enabled prosthetic hand to demonstrate dexterous manipulation with efficient, and compliant actuation.