Design of Finite Time Settling Regulators for Linear Systems

Abstract

Earlier development of finite time settling controllers focused on the structure of the control law which consists of feedback and feedforward parts. In this structure, the feedback part is designed separately to satisfy certain performance specifications in the frequency and/or the time domain. The feedforward part is determined from the feedback control law, and therefore, there exists one-way coupling of both parts. In this paper, we propose a modification in the control structure that enables the designer to regulate the bounds of the control input and the state responses. We show that the finite time settling control problem can be transformed into a linear quadratic regulator one. This transformation results in a two-way coupling of the feedback and the feedforward control laws. We verify that the robustness property of the control strategy is preserved despite its structural change. In addition, we give guidelines for the selection of the feedforward control law.