Design of Mechanism and Control Strategies for Assistive Systems to Remedy the Decrease in Physical Strength of the Aged and the Disabled

许多统计数据表明，人口老龄化日益严重，老年人的体力能力也在下降。本研究将探索辅助外骨骼的控制策略，以帮助弥补老年人上肢力量的丧失，使这一日益增长的人群（以及体弱者和残疾人）过上更加自给自足、富有成效的生活。大多数人类辅助应用程序涉及不同的阶段，每个阶段都有自己独特的动态和这些阶段之间的某些/不确定的过渡。作为本研究主题的设备将被设计为允许主动改变机械动态特性。此外，它将与适应每个阶段的智能控制策略集成，以确保组合人-装置系统的性能和安全性。这项研究的结果将改变辅助设备的行为方式。这项研究涉及多个学科，包括机械设计、控制理论和生物工程。多学科的性质将有助于发展研究参与者的领导技能，并对工程教育产生积极影响。人机辅助系统的设计与控制的集成是本研究的重点。控制问题将从混合系统的角度来阐述。该研究包括人类辅助系统的动态分割，过渡的识别，以及适合每个部分的控制结构的发展。通过这项研究，以下知识将被转化或推进：1)人类肌肉骨骼建模，混合系统建模，以及典型上肢任务中可变动力学的模式（相位）识别；2)具有可变动态特性的辅助装置的设计原则，适用于模式变化的感知和反应，包括脉冲动态变化；3)辅助系统混合切换策略（保护）的合成与分析的原理与方法，以确保用户的安全。混合系统理论与实际辅助系统的整合将是推进我们对这种物理人机交互动力学理解的关键。

Many statistics show a growing aging population, as well as a decrease in the strength capabilities of individuals comprising the aging population. This research will explore control strategies for assistive exoskeletons to help remedy the loss of strength in aging individuals' upper-extremities, allowing this growing segment of the population (as well as the infirm and handicapped) to live more self-sufficient, productive lives. Most human assistive applications involve distinct phases, each with their own unique dynamics and certain/uncertain transitions between these phases. The device that is the subject of this research will be designed to allow for active changing of the mechanical dynamic characteristics. In addition, it will be integrated with an intelligent control strategy that adapts to each phase in order to ensure the performance and safety of the combined human-device system. Results from this research will transform the way assistive devices behave. This research involves several disciplines including mechanical design, control theory, and bioengineering. The multi-disciplinary nature will help develop leadership skills of the participants in the research and positively impact engineering education.Integration of design and control of human assistive systems is emphasized in this research. The control problem will be formulated from the viewpoint of a hybrid system. The research encompasses the segmentation of the dynamics of the human assistive system, the identification of transitions, and the development of control structures appropriate for each segment. With this research, the following knowledge will be transformed or advanced: 1) human musculoskeletal modeling, hybrid system modeling, and mode (phase) identification of variable dynamics in typical upper-extremity tasks; 2) design principles for assistive devices with variable dynamic characteristics suitable for sensing of and reactions to mode changes, including impulse-like dynamic changes; and 3) principles and means for synthesis and analysis of hybrid switching policies (guards) for assistive systems to ensure safety of the user. The integration of the hybrid system theory with practical assistive systems will be the key to advancing our understanding of dynamics in such physical human-robot interactions.