Extended State Observer-Based Estimation for Nonlinear Markov Jump Systems With Dynamic Event-Triggered Communication

This paper primarily investigates the extended state observer design issue for nonlinear Markov jump systems under limited communication bandwidth. In order to achieve reasonable network resource configurations while estimating the system states and total disturbances, a continuous-time dynamic event-triggered strategy is incorporated into the observer design. Accordingly, the transmission of the measured output occurs only when the predefined conditions are satisfied. In contrast to the conventional static triggering strategy, the dynamic event-triggered mechanism enables adaptive threshold adjustment by introducing an internal dynamic variable. In addition, a carefully constructed intermediate variable is considered to indirectly obtain the estimate of the extended state. Then, combining Lyapunov method with stochastic analysis technique, it is proved that the estimation error is exponentially ultimately bounded in mean square under the proposal. At last, a case study of the mass-spring system with four operating modes is presented to illustrate the feasibility of the proposed design approach.