SMC for Networked Non-Homogeneous Hidden Semi-Markov Switching Systems With Cyber Attacks and Application to DC-DC Buck Converter Circuit

Abstract

This study addresses the problem of sliding mode control (SMC) for networked non-homogeneous hidden semi-Markov switching systems under semi-Markov kernel and cyber attacks, in which the limited dwell time information is related to the probability density function. Firstly, the model of networked non-homogeneous hidden semi-Markov switching systems under cyber attacks is constructed. Considering that it is difficult to obtain all the real modes in actual systems, a hidden semi-Markov chain is utilised to determine the hidden mode switching of the underlying system, which is more realistic than semi-Markov chain. Based on restricted dwell time probability density function, the common assumption is relaxed under completely known probability density function. The main innovation is to build a suitable SMC scheme under denial-of-service attacks to achieve quasi-sliding mode that eliminates the effect of uncertain parameters. By means of the Lyapunov function depending on the system mode and the elapsed time, stability criteria are considered for the corresponding model. Finally, a DC-DC buck converter circuit model is introduced to validate the practicality of the proposed strategy. Note to Practitioners—Note that the control community has witnessed a tremendous development in networked control systems with wide applications in many practical models, enabling remote control in a sensitive manner through communication networks. Although networked control systems offer obvious advantages, they always suffer from particular difficulties associated with network transmission, such as cyber attacks and packet losses. With the rapidly developing network and communication technologies, networked non-homogeneous hidden semi-Markov switching systems have attracted much attention due to strong capabilities in modeling dynamical systems with abrupt changes in structures or parameters. It is worth emphasising that it is difficult to obtain all real mode information of networked semi-Markov switching systems. This paper offers a novel approach for researchers to investigate the SMC for networked non-homogeneous hidden semi-Markov switching systems under semi-Markov kernel and cyber attacks.