Average consensus of multi-agent system with binary-valued observations

This paper considers average consensus problem of a multi-agent system with binary-valued observations. The information network transfers binary-valued information and the control is a quantity of transportation between neighbors. We propose a two-scale average consensus algorithm with a constant gain and prove the convergence and convergence rate of the algorithm. In the first stage, each agent estimates the states of its neighbours based on binary-valued observations for a holding time. In the second stage, the system updates the states of agents by a control law which is designed as a quantity of transportation between neighbors with a constant gain. The algorithm is proved to be convergent and asymptotically efficient, meanwhile, the renewal of agents can lead to average consensus. Finally, simulations illustrate the effectiveness of the obtained results.