Event-based resilient filtering for stochastic nonlinear systems via innovation constraints

This paper considers the problem of event-based resilient filtering for a class of stochastic nonlinear systems subject to the impact of outliers. The transmitted data governed by an event-based communication protocol could suffer from malicious attacks due mainly to the network unreliability, which gives rise to the phenomena of outliers or abnormal values. A factitious saturation constraint on innovations is carried out to remove these abnormal data in the designed filter in order to improve the filtering reliability. Furthermore, a gain variation is also taken into account to realize the resilient requirement of the designed filtering scheme. By virtue of the Lyapunov stability theory, a sufficient condition is derived to check the ultimate boundedness of filtering error dynamics in mean-square sense. Furthermore, an analytic formula of the desired filter gain and the ultimate bound of filtering errors are obtained through the utilization of matrix inequality techniques. Finally, some simulation results are provided to illustrate the superiority of the developed filtering scheme . (c) 2020 Elsevier Inc. All rights reserved.

Automated summary

This paper addresses the event-based resilient filtering for stochastic nonlinear systems affected by outliers, utilizing gain variation and a factitious saturation constraint to improve reliability. By employing Lyapunov stability theory and matrix inequality techniques, sufficient conditions for ultimate boundedness of filtering error dynamics in mean-square sense are derived. The study also provides analytical formulas for desired filter gain and ultimate bound of filtering errors, along with simulation results illustrating the superiority of the developed filtering scheme.