Active Robust Fault Detection of Closed-Loop Systems: General Cost Cas

Abstract

Active failure detection consists of finding an input signal (auxiliary signal) such that its use allows perfect detection of faults in a multi-model framework. In this paper we consider the problem of optimal auxiliary signal design for a linear uncertain system controlled by a linear feedback. The optimization criterion used is a worst case quadratic cost, which in practice is the same cost used for the design of the controller. Even though we suppose that the feedback controller is given, the methodology developed here should be useful for feedback design to achieve both control and fault detection objectives.