Robust neuro-adaptive bipartite formation control for nonlinear multi-agent systems subject to Markovian switching topologies and additive noise

Abstract

This article explores the problem of robust neuro-adaptive bipartite time-varying formation (BTVF) control for nonlinear multi-agent systems (MASs) under Markovian randomly switching topologies and additive noise. Both leaderless and leader-following cases are considered, with the communication topology modeled as a continuous-time Markov chain. Additionally, the impact of communication noise in realistic scenarios is addressed. To mitigate disturbances and nonlinearities, an adaptive control strategy incorporating neural network (NN) approximation is employed. With the aid of the infinitesimal generator and the indicator function, the expected leaderless and leader-following BTVF can be achieved. Towards the end, the proposed theoretical method is confirmed through two numerical examples.