Fully Distributed Fuzzy Adaptive Consensus for Nonlinear MASs With an Active Leader: State and Output Feedback Designs

This article investigates the fully distributed fuzzy adaptive consensus problem for nonlinear multi-agent systems (MASs) with an active leader. The active leader can be controlled in a real-time manner by an outside operator, and the control signals of the leader and its upper bound are completely unknown to all followers. The problem is studied in both state feedback and output feedback cases. In the case where the system states are available, a fuzzy adaptive controller that can dynamically compensate for the approximation errors of fuzzy logic systems (FLSs) is designed for the first time. This reduces the high demand for fuzzy approximation accuracy while maintaining the control performance. In another case where only the disturbed system output is available, an augmented observer that can handle arbitrary unknown measurement disturbances is novelty constructed. In both cases, by introducing adaptive auxiliary functions, unknown leader inputs are effectively disposed of, and the tracking performance of the system is nicely guaranteed. Finally, simulation examples are given to illustrate the feasibility and advantages of the developed strategies.