A Graph Neural Network Based Multi Agent System for Traffic Engineering

Group formation control problems for nonlinear second-order multi-agent systems are considered using neural networks. Multiple agents are divided into subgroups to form specified sub-formation and construct interactions among subgroups based on graph theory. First of all, an adaptive protocol is designed for nonlinear second-order multi-agent systems based on neural networks. Then sufficient conditions for the multi-agent systems to achieve group formation are given. Finally, a numerical example is provided to verify the proposed conditions.