H∞ optimal control of a minimal state space realization model of coordinated fingers movement

Abstract

A model can be greatly simplified in terms of analysis if we work with a lower order model instead of higher order models. This simplification leads to less complicated control design. In this paper we analyse a minimal realization biomechanical model of reflexive movement of human finger when the little finger is bent. This is a 6th order minimal state space realization model for movement simulation of two fingers. We developed and compared different results of an H∞ optimal controller to achieve a stable response. Exogenous inputs in terms of parametric uncertainties were added to the model. We manipulated the values to get a better controller. Furthermore, we minimized the effects of the disturbances and the impact of the measurement noise and modeling errors by tuning the controller for optimum performance. We simulated the model in MATLAB/Simulink for study of stability analysis and performance comparison. Our simulation results demonstrate the applicability of the model towards active prosthetic, and help to understand central nervous system in physiologically coordinated movement and constraint on motor.