Adaptive event-triggered filtering for semi-Markov jump systems under communication constraints

Abstract

This paper investigates the problem of adaptive event-triggered filtering for a class of semi-Markov jump systems. A more general scenario of semi-Markov jump systems under communication constraints is explored, where the output signal occurs with saturation nonlinearity, packet loss, and quantization. The saturation nonlinearity and packet dropout are described by two random variables subject to Bernoulli distribution. The mode-dependent adaptive event-triggered mechanism is designed by combining the system mode information. Based on the output error, the trigger threshold adaptive law is designed to make the threshold change dynamically. The mode-dependent time-varying filter is constructed and sufficient conditions for the desired filter with H∞ performance are obtained. Finally, a single-link robot arm system verifies the effectiveness of the proposed filter design method, and the adaptive event-triggered protocol can effectively reduce the number of data transmissions.