© 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting /republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works

In this study, actuator fault diagnosis of singular delayed linear parameter varying (SDLPV) systems is considered. The considered system has a time-varying state delay and its matrices are dependent on some parameters that are measurable online. It is assumed that the measured parameters are inexact due to the existence of noise in real situations. The system with inexact measured parameters is converted to an uncertain system. Actuator fault diagnosis is carried out based on fault size estimation. For this purpose, the system is transformed to a polytopic representation and then a polytopic proportional integral unknown input observer (PI-UIO) is designed. The proposed observer provides simultaneous state and actuator fault estimation while attenuating, in the H8H8 sense, the effects of input disturbance, output noise and the uncertainty caused by inexact measured parameters. The design procedure of PI-UIO is formulated as a convex optimisation problem with a set of Linear Matrix Inequality (LMI) constraints in the vertices of the parameter domain, guaranteeing robust exponential convergence of the PI-UIO. The efficiency of the proposed method is illustrated with an electrical circuit example modelled as an SDLPV system.

Peer Reviewed