New H-infinity observer-based control for delayed LPV stochastic system

Abstract

An H-infinity observer-based control problem of delayed Linear Parameter Varying (LPV) stochastic system is addressed in this paper. In the LPV stochastic system, stochastic behaviors and external disturbance in real environment are respectively described by stochastic difference function. Furthermore, the delay effects on state and control input are considered as time-varying case for wide application. For the control problem, a lemma dealing with the delay terms is proposed to increase some slack variables. Based on the proposed lemma and the chosen Lyapunov-Krasovskii function, some sufficient conditions are derived to relax the conservatism of stability criterion. Besides, disturbance effect on state and estimation error is constrained via H-infinity performance for improving transient response. Through using the extended projection lemma, the derived conditions are formulated as strict Linear Matrix Inequality (LMI) problem. To avoid the potential oscillation, another lemma is proposed to separately assign the performance level during solving the conditions. All feasible solutions can be directly obtained by convex optimization algorithm without using extra calculation. Therefore, an observer-based controller is established to guarantee the asymptotical stability and H-infinity performance in the mean square. Finally, some numerical simulations are provided to demonstrate the effectiveness and applicability of the proposed method.