Switching Control for Parameter Identifiability of Uncertain Systems

TL;DR

This paper demonstrates that switching among a finite set of linear controllers can ensure parameter identifiability in uncertain linear systems, with implications for fault detection and adaptive control.

Contribution

It introduces a method using time-varying controllers to achieve parameter identifiability even with infinite uncertainty sets.

Findings

Parameter identifiability can be generically achieved through switching.

Finite controllers suffice for systems with non-finite uncertainty sets.

Implications extend to fault detection, isolation, and adaptive control.

Abstract

This paper considers the problem of identifying the parameters of an uncertain linear system by means of feedback control. The problem is approached by considering time-varying controllers. It is shown that even when the uncertainty set is not finite, parameter identifiability can be generically ensured by switching among a finite number of linear time-invariant controllers. The results are shown to have several implications, ranging from fault detection and isolation to adaptive and supervisory control. Practical aspects of the problem are also discussed in details.