Robust consensus of uncertain multi agent systems with one‐sided Lipschitz nonlinearity

Abstract

This article develops protocols for the consensus of an uncertain nonlinear multi-agent system that satisfies the one-sided Lipschitz and quadratic-inner bounded conditions. Fully distributed consensus protocols with observer-based and observer-less schemes are investigated. For the observer-based system, a full order Luenberger observer is designed. To overcome the difficulties of designing the controller and observer-based controller gains, sufficient conditions are provided with a complete linear matrix inequality characterization. Finally, two simulation examples are provided to illustrate the effectiveness of the proposed approach.

Fixed‐time sliding mode consensus tracking for second‐order nonlinear multi‐agent system with persistent dwell‐time switching topologies

“…Recently, researchers have increasingly focused on achieving consensus of MAS in finite time. While traditional analyses of asymptotic stability [21][22][23] hold theoretical significance, practical systems demand methods that operate within finite time frames. Therefore, there has been a significant increase of research in developing finite-time consensus in recent years, 24,25 which can lead to faster convergence, better tracking accuracy, and increased robustness.…”

Iterative learning consensus control for one‐sided Lipschitz multi‐agent systems

“…By using the reinforcement learning method, the data‐driven optimal consensus was considered in Reference 8 for discrete‐time MASs. In Reference 9, the constrained optimal coordination problem was discussed for nonlinear MASs. For high‐order nonlinear systems (NSs), the backstepping‐based adaptive consensus control problem was addressed in Reference 10.…”

“…Moreover, the finite‐time consensus problems for MASs were also studied (see References 11‐13 and references therein). Note that the consensus in References 4‐13 can be ensured only when time tends to infinity or after a finite‐time, which means the consensus tracking on the whole time interval can't be reached exactly.…”