EAGER: A Holistic Approach to Adaptive Robust Control of Systems with Uncertain Nonsmooth Nonlinearities with Application to Cable-Conduit Actuated Surgical Robotic Devices

点击翻译按钮获取中文摘要

The research objectives of this EArly-concept Grants for Exploratory Research (EAGER) award are to (i) develop effective compensations for typical nonsmooth nonlinearities to maximize the achievable control performance in implementation without the sensitivity problems of existing perfect inversion based designs; (ii) develop quantitative parameter estimation algorithms for uncertain nonsmooth nonlinearities through a combination of explicit monitoring of experimental conditions and the use of physical model based parameter estimations; (iii) take a holistic mechatronic approach to the control of systems having nonsmooth nonlinearities to develop suitable model structures for controls and effective adaptive robust controllers that can achieve high performance in practice without being overly complex or having degraded life-expectancy of system components; and (iv) apply the proposed holistic mechatronic control framework to the control of cable-conduit actuated mechanisms for surgical robotic systems, the ultra-precision control of piezoelectric actuator driven nano-positioning stages for high-speed nano-technology operations, and the control of other modern mechanical systems such as high-speed electro-magnetic motor driven and electro-hydraulic/pneumatic actuated ones for an improved productivity and quality.The proposed research departs from the prevailing design philosophy of perfect inversion based nonsmooth nonlinearity compensation and provides a novel perspective to the effective compensation of nonsmooth nonlinearities. In addition, as opposed to focusing on only a subset of an entire control problem to be dealt with in reality, the proposed research takes a holistic mechatronic approach to the modeling and control of systems with nonsmooth nonlinearities. The set of tools generated through the research will enable engineers to systematically quantify the effects of various implementation imperfections to maximize the achievable control performance in practice without the sensitivity problem of existing designs. If successful, the results of this research will be integrated into the curriculum to have a lasting impact on future control education. The results will also provide a solid foundation to the design of a new generation of controllers which help industry build modern machines of great performance and intelligence with built-in machine component health monitoring capability.