Fuzzy Adaptive Resilient Output Control for Nonlinear Multi-Agent Systems Under Byzantine Agents

Abstract

This article studies the fuzzy adaptive resilient output control for nonlinear multi-agent systems (MASs) under Byzantine attacks. Firstly, fuzzy logic systems (FLSs) are adopted to approximate the unknown nonlinear dynamics, then a fuzzy state observer is constructed to identify the unmeasured states in the controlled system. The mean-subsequence-reduced (MSR) technique is introduced to filter the unavailable neighbor information of each agent. On the basis of the distributed backstepping technique, a special Nussbaum function is utilized to construct an attack compensation scheme. Combining the compensation scheme and MSR technique, an observer-based robust fuzzy adaptive resilient output control design algorithm is proposed. The developed resilient control method enables the system outputs follow a given protocol under Byzantine attacks. Finally, the feasibility and effectiveness of the proposed control method and theory can be illustrated via the simulation results. Note to Practitioners—In this article, an observer-based resilient fuzzy adaptive output control issue is studied for nonlinear multi-agent systems subjected to Byzantine attacks. Fuzzy state observer is constructed to identify the unmeasured states. Since there exist the Byzantine attacks in the controlled system, the neighbor information of each agent cannot be fully used in the controller design. Thus, compared with the existing related results, the mean-subsequence-reduced technique can be introduced in this article to filter the unavailable neighbor information of each agent. Besides, an attack compensation control method via a special Nussbaum function is designed to resist the effect of uncertainties and malicious data caused by attacks. Simulation results ensure the feasibility of the developed resilient control approach and theory.