Fuzzy T-S Model Based Fault Tolerant Control Using Immeasurable Decision Variables

Abstract

Developing a fault-tolerant control strategy is the goal of this work so as to reduce the gap between the dynamics of faulty system and reference model. The considered nonlinear systems (NLS) are represented by fuzzy T-S models using immeasurable decision variables in the presence of faults affecting simultaneously the actuator and the output. First, a Proportional Integral (PI) multiobserver is designed to estimate the faults and the faulty system states. Then, a Fault Tolerant Tracking Control (FTTC) is proposed to compensate the impacts of faults and ensure the trajectory dynamics tracking of the reference model regardless of the existence of faults. The performance index using the L2 norm approach is taken into consideration so as to ensure the minimization of the disturbance impacts not only on the tracking error but also on the estimation error of the faulty system states. Finally, an academic example is conveyed to prove the validity, efficiency and advantages of the suggested FTTC approach.