Antagonistic Interactions-Based Adaptive Event-Triggered Bipartite Consensus Quantized Control for Stochastic Multiagent Systems

Abstract

In a network of multiagent systems (MASs), the interaction between agents is not necessarily cooperative, but may be antagonistic, when the communication topology of a fixed spanning tree is formed. Although the cooperative control methods have been widely studied, the control methods aiming to the coexistence of antagonistic and cooperative interactions are rarely researched. Therefore, for antagonistic and cooperative interactions, this article addresses the bipartite cooperative quantized control problem for nonlinear stochastic MASs by using an event-triggered mechanism. The MASs model we chose allows cooperative and antagonistic interactions to coexist, with every agent being an independent individual. The Gaussian function is introduced to simulate cooperation and competition between two agents. Fuzzy neural networks are applied to approximate the unknown nonlinear functions, which reduces the burden of online calculation. Furthermore, to reduce the communication burden, an adaptive event-triggered approach with a varying threshold is developed. Meanwhile, according to sector bound property, a nonlinearity decomposition strategy of asymmetric hysteresis quantizer is applied. Compared with the previous situation that only event-triggered mechanism or quantizer is considered, this article reduces the communication burden to the greatest extent by codesigning the event-triggered mechanism and quantizer. Based on the backstepping technique, we propose a novel distributed adaptive control protocol, which not only achieves bipartite consensus control, but also guarantees that all the signals in the closed-loop system are bounded in probability, and the bipartite synchronization error converges to a small range near the origin. Finally, the performance of the proposed control strategy is illustrated through simulation.