ShiLei Zhao, TieJun Zhang, and YuPeng Liu
Iterative learning, Fault-tolerant control, descriptor time delaysystem, unknown input observer
Based on the iterative learning (IL) strategy and unknown input
observer (UIO), the problem of actuator fault estimation and
fault-tolerant control design for linear systems is studied in this
paper. The aim of employing the UIO is to estimate system states
accurately for feedback control with the eﬀects of disturbances.
The IL update law is adopted to learn the fault by using the
previous fault information and the UIO output estimation error.
The IL update law which can be referred to time-delay control,
the feedback input error, the system state and the state estimation
error is organized into a kind of time-delay descriptor system, and
the stability of this descriptor system is easily solved in terms of
linear matrix inequalities, Lyapunov–Krasovskii stability theory and
H∞ technique. Finally, a numerical example has been studied to
illustrate the eﬃciency of the proposed approach.