Robust H∞ filtering of discrete-time nonhomogeneous Markovian jump systems with dual-layer operation modes

This paper investigates the robust H-infinity filtering problem for discrete-time nonhomogeneous Markovian jump systems (MJSs) with dual-layer operation modes: system operation (low-layer) mode and event (high-layer) mode. In the configuration of dual-layer modes, the event mode is used to represent a driving factor that influence the distribution of transition probabilities (TPs) of the system operation mode, which ultimately allow to form locally distributed TP boundary sets (TPBSs) from a single TPBS. Based on the formation of local TPBSs, which has a noticeable effect on removing redundant resources of constraints imposed on TPs, this paper first proposes a proper relaxation technique capable of turning double-parameterized matrix inequalities into linear matrix inequalities (LMIs). Then, with the aid of the proposed relaxation technique, a less conservative condition filter design is derived in terms of LMIs by using a new augmented parameter matrix that assists the one-step relaxation process. (C) 2018 The Franklin Institute. Published by Elsevier Ltd. All rights reserved.