A Method of Indirect Configuration Propagation With Estimation of System State in Networked Multi-Agent Dynamic Systems

Networked multi-agent dynamic system (NMDS) exhibits emergent swarming behavior by using local control laws based on spatial information from the nearby environment and adjacent agents. One of the challenges in NMDS is the lack of appropriate methods for operators to interact with a large group of agents. Studies on human-agent interaction have mainly focused on the direct influence on the behavior of the multi-agents, but it is not feasible for a large-scale system in an error-prone environment. In this letter, a robust method for indirect configuration propagation by a human operator is proposed to influence the swarming behavior of NMDS in a decentralized manner. In addition, an estimation method is proposed for the operator to observe the system state, which helps the operator provide an optimal input to the NMDS at the appropriate time. Finally, we demonstrate the superiority of the proposed method through theoretical analysis and simulation results.