Principles of Fault Tolerant Control

Abstract

A closed-loop control system in an engineering process may have unsatisfactory performance or even instability when faults occur in actuators, sensors or other process components. Fault-tolerant control (FTC) involves the development and design of special controllers that are capable of tolerating the faults whilst still maintaining desirable and robust performance and stability properties. FTC designs involve knowledge of the nature and/or occurrence of faults in the closed-loop system either implicitly or explicitly using methods of fault detection and isolation (FDI), fault detection and diagnosis (FDD) or fault estimation (FE). Here we replace the abbreviation FDD by the more familiar FDI. FTC controllers are reconfigured or restructured using FDI information so that the effects of the faults are reduced or eliminated within each feedback loop in active or passive approaches or compensated in each control-loop using FE methods. The abbreviations FDI and FDD are often used inter-changeably according to the varied usage in different Control/Aerospace communities. In this article a non-mathematical outline of the essential features of FTC systems is given with important definitions, including a classification of FTC systems into either active/passive model-based approaches, with examples of some well-known strategies as well as important application areas.