Robust Stochastic Sampled-data-based Output Consensus of Heterogeneous Multi-agent Systems Subject to Random DoS Attack: A Markovian Jumping System Approach

• Published in: International journal of control, automation, and systems 2019, Automation, and Systems, 17(7), , pp.1687-1698
• Main Authors: Ni, Hongjie, Xu, Zhenhua, Cheng, Jun, Zhang, Dan
• Format: Article
• Language: English
• Subjects: Automation; Communication channels; Communications systems; Control; Control algorithms; Control systems design; Controllers; Decoupling; Denial of service attacks; Design; Discrete time systems; Engineering; H-infinity control; Jumping; Markov chains; Mathematical analysis; Matrix methods; Mechatronics; Multiagent systems; Output feedback; Robotics; Sampling; Systems design; Tracking problem; 제어계측공학
• ISSN: 1598-6446; 2005-4092
• DOI: 10.1007/s12555-018-0658-9

This paper addresses the robust stochastic sampled-data-based output feedback (SSDBOF) consensus controller design for a network of continuous-time heterogeneous multi-agent systems (MASs) in the presence of denial-of-service (DoS) attack. Under the mild assumption that the sampling instant is stochastically triggered and satisfies the Markov property, a homogeneous Markovian jump system (MJS) method is introduced that is capable of modeling the stochastic sampled-data-based control system. Furthermore, the randomly occurring Deny-of-Service (DoS) attack problem is also taken into account due to the existence of potential adversary that tries to block the communication channels. A novel discrete-time stochastic Markovian system model is first introduced that enables us to deal with the stochastic sampling and random DoS attack phenomena in a unified framework. Then by adopting the decoupling scheme, some sufficient conditions are proposed such that all the outputs of the following agents can track the output of the leading agent, and the prescribed H ∞ performance level is also guaranteed. In our work, the SSDBOF consensus controller design method is transformed to a feasibility problem subject to the linear matrix inequality (LMI) constraints. The theoretical results are finally applied to solve the position tracking problem of a network of vehicle systems.