Event-Triggered Neural Network-Based Control for Prescribed-Time Cluster Quasi-Consensus in Multi-Agent Systems
Shi Qiu, Wenyan Tang, Rui Shu · 2025
This paper addresses the prescribed-time cluster quasi-consensus problem for unknown nonlinear multi-agent systems (MASs) with a directed communication topology. A radial basis function neural network (RBFNN)-based adaptive control protocol is proposed to handle the lack of precise system dynamics. A distributed control strategy ensures agents reach their cluster consensus values within the prescribed time, and an event-triggered mechanism reduces communication costs without sacrificing accuracy. Theoretical analysis confirms both strategies achieve prescribed-time cluster quasi-consensus and avoid Zeno behavior. Simulation results validate the approach, demonstrating its efficiency in reducing communication and computation costs.