Predictive consensus tracking of multi-agent systems in the presence of Byzantine agents and connection loss of reference signals

This article proposes the consensus tracking of multi-agent systems, in the presence of Byzantine agents and loss of connection between agents and the reference signal. The main purpose of this work is to steer a multi-agent system to the desired reference, assuming that there are some agents which send incorrect information to the other agents, and the connection between the agents and the reference signal may lose. Using graph properties alongside the weighted-mean-subsequence-reduced (W-MSR) algorithm led to a novel control method for multi-agent systems. With the aid of model predictive control (MPC), the agents reach the desired consensus with connection loss of reference signals, misbehaving agents in the system, and constraints on the input signals. Combining MPC with the W-MSR algorithm results in an algorithm called reference-based-W-MSR (RBW-MSR), which has made the consensus point fixed according to a predetermined reference. A theorem is illustrated to guarantee the consensus with the interruption of reference signals and agents. The effectiveness of the proposed algorithm is illustrated via simulation results.