mathrm \infty Consensus PID Control of Multi-Agent Systems with Stochastic Communication Protocols Under Network Attacks

• Published in: Chinese Control and Decision Conference pp. 5898 - 5905
• Main Authors: Jinbo, Song, Junning, Liu, Shiqiang, Zhu, Mengdi, Chang, Chao, Xu
• Format: Conference Proceeding
• Language: English
• Published: IEEE 16.05.2025
• Subjects: Consensus control; Control theory; cyber attack; Cyberattack; H_{\infty} consensus control; mean-square exponential stability; Multi-agent systems; Protocols; Schedules; Security; Stability; stochastic communication protocol(SCP); Sufficient conditions; Time-varying systems
• ISSN: 1948-9447
• DOI: 10.1109/CCDC65474.2025.11090460

This paper is concerned with the observer based H_{\infty} consensus control problem for a class of discrete-time multi-agent systems(MAS). The stochastic communication protocol (SCP) obeying Markovian chain is adopted to schedule the information transmission among agents, by which only one neighbor agent is stochastically selected and the information transmitted to the agent at each transmission instant, so that to avoid the problem of channel congestion and enhance communication security. To comprehensively depict the communication scheduling of the MAS, multiple Markov chains are mapped. Furthermore, the cyber-attacks, which include both denial-of-service and deception attacks, are allowed to be randomly occurring as regulated by two sequences of Bernoulli distributed random variables with certain probabilities. Then, a novel observer-based proportional-integralderivative (PID) controller is proposed such that the closedloop system achieves the desired H_{\infty} consensus performance. Sufficient conditions are derived under which the mean-square exponential stability is guaranteed. The PID controller gains are obtained by solving a set of linear matrix inequalities. Finally, the validity of the developed design approach is verified via an illustrative example.