Fast Consensus of Second-Order Multi-agent Systems Based on Past State Prediction

In this paper, fast stationary and dynamic consensus protocols are proposed for multi-agent systems (MASs) with undirected fixed topologies by predicting the state information in the past. The system stability is analyzed from the frequency domain analysis by virtue of algebraic theory. It is proved that the MAS using past-state prediction reaches consensus faster than its counterpart merely using current state information when the multi-agents’ past state information value and prediction strength value are properly selected. The necessary and sufficient conditions on the stability of MASs and the sufficient conditions on MASs’ convergence speed are established. Finally, the effectiveness of the proposed consensus protocol is verified through numerical simulations.