A sampled-data approach to distributed H-infinity resilient state estimation for a class of nonlinear time-delay systems over sensor networks

摘要

This paper deals with the distributed sampled-data H-infinity state estimation problem for a class of continuous-time nonlinear systems with infinite-distributed delays. To cater for possible implementation errors, the estimator gain is allowed to have certain bounded parameter variations. A sensor network is deployed to acquire the plant output by collaborating with their neighbors according to a given network topology. The individually sampled sensor measurement is transmitted to the corresponding estimator through a digital communication channel. By utilizing the input delay approach, the effect of the sam-pling intervals is transformed into an equivalent bounded time-varying delay. A set of sampled-data distributed estimators is designed for the addressed nonlinear systems in order to meet the following three performance requirements: (1) the asymptotic convergence of the estimation error dynamics; (2) the H-infinity disturbance attenuation/rejection behavior against the exogenous disturbances; and (3) the re-silience against possible gain variations. A Lyapunov functional approach is put forward to obtain the existence conditions for the desired estimators which are then parameterized in light of the feasibility