Asynchronous Secure Control of Fuzzy MJSs With Time Delays: A Multi-Level Probabilistic Event-Triggered Mechanism

This paper investigates the event-triggered secure control problem for fuzzy Markov jump systems (MJSs) with time-varying delays. In order for higher resource efficiency, a multi-level probabilistic event-triggered mechanism (MLPETM) is newly established, which reduces the data transmission frequency by investigating possible variations of triggering thresholds. Given the network-induced phenomena, the MLPETM’s output may undergo cyber-attacks, and the controller may not switch synchronously along with the system. Asynchronous secure controllers are thus developed to perform secure control of fuzzy MJSs. Sufficient conditions dependent on fuzzy basis functions, state delays and jumping modes are established to guarantee the dissipative finite-time boundedness of the resultant closed-loop system. Afterward, an easy-to-check fuzzy secure control algorithm is developed to compute the controller gains. Finally, the one-link robotic arm system and the truck-trailer system are borrowed to validate the proposed results.
Note to Practitioners
—Networked control systems (NCSs) play a key role in real applications like automobile industry, power systems, and autonomous mobile robots. As a key part, the communication network such as Controller Area Network is embedded into the NCSs to perform the communication tasks between different control units. In spite of great benefits from the communication network, both the security and the efficiency issues of the data transmission are challenging for the engineering designer. Besides, the target physical system is usually nonlinear, and is very likely to undergo abrupt changes in the networked environment. In terms of these challenges, this paper studies the asynchronous security control of networked Markov jump systems with time-varying delays and randomly occurring cyber-attacks. A multi-level probabilistic event-triggered mechanism (MLPETM) is newly established, which studies the effect of the stochastic variation regarding the triggering threshold on the triggering performance. Potential cyber-attacks are modelled in a randomly occurring fashion and then absorbed into to the control law which allows operation modes asynchronous to the system modes. Finally, the one-link robotic arm system and the truck-trailer system are borrowed to check the proposed findings. Note that both the transmission security and communication traffic regulation are well concerned, which renders the developed method applicable to practical systems.