Fault estimation in discrete-time linear systems with mixed uncertainties using proportional integral observer

Alexandria Engineering Journal

Rosmiwati Mohd-Mokhtar¹  Masood Ahmad² 

[1] Department of Electrical and Computer Engineering, COMSATS University, Islamabad (Lahore Campus), Lahore, Pakistan;School of Electrical and Electronic Engineering, Engineering Campus, Universiti Sains Malaysia, 14300 Nibong Tebal, Pulau Pinang, Malaysia;

关键词: Fault estimation;    PI observer;    Uncertain systems;    Norm;    Robustness;    LMI;   

DOI  :  

来源: DOAJ

【 摘 要 】

In this article, the problem of fault estimation is addressed for the class of discrete-time linear systems subject to mixed uncertainties (norm bounded and stochastic uncertainty) and unknown input disturbances. A robust proportional integral (PI) observer is proposed based on H∞ norm minimization for dealing with uncertainties and disturbances. An augmented observer is constructed first where the sensor fault is treated as a state variable and then, H∞ norm is minimized in the linear matrix inequalities (LMI) framework with the application of the Lyapunov function and Young inequalities. The application of the DC motor system and well-known three tank system is given to demonstrate the performance of the proposed observer. Through simulation results, a proposed PI observer has shown the precise estimation of a sensor fault.

【 授权许可】

Unknown