A Generative Virtual Leader Network Framework for Enhanced Synchronization and Disturbance Resistance in Multi-Agent Systems

This paper introduces a novel control method for multi-agent systems by integrating a Generative Virtual Leader Network (GVLN) to improve synchronization speed and disturbance resistance. Unlike traditional single-layer complex network-based methods, the proposed approach leverages a multi-layer synchronization framework, where the GVLN provides auxiliary control to the follower network. Simulation results demonstrate that this method significantly reduces synchronization time and system oscillations while enhancing robustness against disturbances. Experiments on chaotic systems confirm the superiority of the proposed approach in achieving faster state coordination and improved stability. These findings offer new insights into complex network synchronization and practical multi-agent system control.