From fault diagnosis to reconfigurable control: A unified concept

Abstract

Active fault-tolerant control is generally based on two steps: At first, a diagnostic unit evaluates the measured input and output signals of the plant in order to detect and identify a fault or component failures. At second, after a fault or failure has been found, the controller is reconfigured so as to ensure that the closed-loop system including the faulty plant satisfies again the performance requirements. In the past, most of the authors have considered either the diagnostic step or the reconfiguration step and only a few results are available concerning the complete concept of fault-tolerant control. This paper presents a unified approach that combines fault diagnosis with control reconfiguration and demonstrates the efficiency of this comprehensive method with a hexrotor experiment.