IDENTIFICATION AND CONTROL OF CRITICAL MECHANICAL RESONANT MODES BEYOND THE NYQUIST FREQUENCY IN HIGH-PERFORMANCE MECHATRONICS SYSTEMS

Abstract

Due to the aliasing effect, system dynamics beyond the Nyquist frequency cannot be identified and controlled using traditional techniques. In this dissertation, several identification and control methods are proposed to solve such problems. In the identification part, a two-step parametric approach is designed to identify mechanical resonances beyond the Nyquist frequency. Furthermore, a general multirate approach without injecting any external persistently exciting signals is proposed to identify system dynamics to the Nyquist frequency and beyond. In the control part, an add-on multirate adaptive control scheme is proposed for compensation of uncertain resonances beyond the Nyquist frequency. By employing the steady state response of the plant output including intersample information, two conditions are provided to reject periodic disturbances to the Nyquist frequency and beyond. Also, a disturbance observer-based multirate control scheme is proposed for rejecting such disturbances. All the methods are validated on the voice coil motor actuator in an HDD servo system.