Modelling and identification of residual generator functions for fault detection and isolation of a small aircraft

Abstract

This paper addresses the design of a fault diagnosis scheme based on linear system frequency domain design. The main innovation of the method consists of integrating in one-step only, the derivation of an analytical model of residual generator functions for the diagnosis of the monitored process with the optimal design of the diagnostic system itself. Special attention will be given to the computation of residual generator functions in order to realise a complete fault diagnosis scheme when additive faults are present, as well as the optimisation of the performance that can be achieved under the presence of disturbance and unknown inputs. The combined modelling and identification approach for the computation of residual generators is exploited for fault detection and isolation of a small aircraft, characterised by a non-linear model, in the presence of disturbance. The results obtained in the simulation of the faulty behaviour of a Piper PA30 are finally reported.