Consensus of multi-agent systems with intrinsic nonlinear dynamics and sampled-data information

This paper is concerned with the problem of consensus in directed networks of multiple agents with intrinsic nonlinear dynamics and sampled-data information. The new protocol is induced from a class of continuous-time linear consensus protocols by implementing data-sampling technique and a zero-order hold circuit. By appropriately constructing a Lyapunov-Krasovskii functional and using Finsler's lemma, it is theoretically proved that consensus with asymptotic time-varying velocities in strongly connected directed networks can be achieved over some suitable sampled-data intervals. Particularly, several feasible linear matrix inequalities are established to estimate the maximal allowable upper bound of sampling intervals. The results are then extended to switching systems with balanced topologies. Finally, some numerical simulations are provided to verify the effectiveness of the theoretical results.