Multiscale Control to Meet the Conflicting Nanoscale Performance Requirements

Abstract

In this paper, we consider the problem of designing a multiscale control for plants with conflicting time-domain performance requirements. These results follow from the conventional optimal proportional-integral (PI) control. Four different design methods are proposed: (1) a controller-switch technique which makes use of employing two different controllers designed to meet two different performances and are switched during the course of operation, (2) an integral-reset scheme, which resets the integral term in the control law when the new reference point is reached, (3) controller-switch and integral-reset schemes put together to take benefits of both of them, (4) a model-following approach that uses a dynamic reference model without increasing the overall dimension of the system. The objective of the last scheme is to make the output of the plant track the output of the model as closely as possible. Stability analyses and a comparison between the performances of these methods are given. All these methods give better performances as compared with conventional control schemes. Block diagrams are given and step responses are obtained to demonstrate the proposed methods. A six degrees-of-freedom (DOFs) magnetically levitated (maglev) stage with a second-order pure-mass model has been used to demonstrate the capabilities of the aforementioned control strategies. These strategies are not plant-specific and may be generalized to any higher-order plant.