Design of Least Order Residual Generators for Fault Detection and Isolation with Application to Monitoring Actuator/Surface Faults for a Boing 747 100/200 Aircraft

Abstract

We address the problem of designing residual generators with least dynamical orders to solve fault detection and isolation problems. The basic design procedure of least order residual generators for fault detection combines a rational nullspace method with minimal dynamical covers based order reduction techniques. This procedure underlies the design of residual generators for fault detection and isolation using a bank of scalar output residual generators. State space representation based numerically reliable algorithms are discussed to solve all underlying computational problems. These algorithms have been implemented as robust numerical software in a dedicated collection of tools to solve fault detection and isolation problems. An application of the proposed techniques and developed new software tools to the actuator fault monitoring for a Boeing 747 100/200 aircraft is presented.