Model Development and Model-Based Control Design for High Performance Nonlinear Smart Systems.

摘要

We have developed a unified energy-based framework for quantifying hysteresis and constitutive nonlinearities inherent to piezoelectric, magnetic, and shape memory compounds which is amenable to inversion and subsequent use as an inverse filter for linear control designs. We have also developed stochastic modeling techniques which quantify the highly complex stiffness properties of ionic polymers in a manner which facilitates design in applications ranging from biological/chemical detection to robotic design for aerospace structures. The control component has focused on the development of robust linear designs exploiting nonlinear filters and fully nonlinear algorithms which incorporate modeled physics directly into the control design. Open loop control experiments have been performed and present investigations are focused on closed loop experimental validation of the control theories.