Impulsive consensus of multi-AUV systems under random packet losses and switching communication topologies

Abstract

This paper studies the impulsive consensus problem for multiple autonomous underwater vehicle (multi-AUV) systems with switching communication topologies and random packet losses. To alleviate the bandwidth limitation of underwater acoustic communication, a periodic impulsive communication strategy is adopted, where neighboring AUVs exchange information only at discrete sampling instants. Packet losses are modeled by a Bernoulli stochastic process. Two distributed impulsive consensus protocols are developed for the cases with and without self-velocity measurements, respectively, while neither protocol requires velocity information from neighboring AUVs. By combining algebraic graph theory with matrix analysis, sufficient consensus conditions are established that explicitly characterize the interplay among the communication topology, sampling period, and controller gains. The sufficient conditions permit consensus under packet loss rates as high as 80% when the sampling intervals and controller gains satisfy the derived bounds. Simulation results illustrate the theoretical analysis and demonstrate that the proposed impulsive strategy achieves substantially faster convergence than a representative event-triggered approach under the same packet-loss conditions. • Impulsive consensus of multi-AUV systems is studied under random packet losses. • Two distributed impulsive protocols eliminate neighboring velocity information exchange. • Consensus conditions reveal the interplay among control gains, sampling period, and topology. • Consensus is guaranteed under Bernoulli packet losses of up to 80%. • Faster convergence is achieved than with a representative event-triggered strategy.