Finite-Iterative Learning Consensus Formation Control for Multi-Agent Systems Under Double-Terminal Switching Topologies

In this paper, the finite-iteration learning consensus formation control issue is studied for a class of multi-agent systems under dual-terminal switching topologies. The completion of consensus formation control requires that each agent can receive control information from the leader indirectly or directly. To relax the restrictions on effective communication relationships among agents, the dual-terminal switching topologies, as a novel communication structure, are proposed for the consensus formation control of multi-agent systems. The dual-terminal switching topologies are composed of switching topologies and iterative-varying topologies corresponding to the variation of communication relations in the time axis and iteration axis, respectively. Therefore, the time axis and iteration axis denote the implication of dual-terminal. The full spanning tree is not required for the proposed topologies at the time terminal, and the missing information among agents can be compensated by the iteration terminal. Based on the communication relationship of dual-terminal switching topologies and the iterative learning control strategy, the corresponding iterative formation error is redefined to perform the desired control task. In order to avoid falling into the infinite iteration of learning, finite-iteration learning control approach is proposed for multi-agent systems with dual-terminal switching topologies. The convergence of formation error is verified by using the contraction-mapping approach. Finally, the effectiveness and availability of the proposed finite-iteration learning-based consensus formation control strategy are measured through a numerical example.