Stability And Stabilization Of Networked Control Systems With Communication Constraints

In recent years, Networked Control System (NCS), as a research focus in automaticcontrol, has attracted more and more attention of researchers from various fields. With therapid development of technology and industrial demand, we have been motivated to dofurther research to provide more precise, more efficient and more humane workingconditions. Feedback Control, Robust Control, Fault-Tolerant Control, Predictive Controland Guaranteed Cost Control were put forward consecutively. However, as an emergingdiscipline, the theoretic research on NCS can be improved greatly. In this thesis, we focuson stability analysis and controller design of NCSs with communication constraints basedon Lyapunov stability theory. The novelties of this thesis lie in:First, in most of the existing studies dealing with stability analysis of NCSs with timedelay and packet dropout, the packet dropout phenomenon is commonly treated as timedelay, then the sufficient conditions for stability of NCSs is derived based on a delay upperbound dependent Lyapunov function. In this thesis, however, the quantitative relationshipbetween packet dropout rate and time delay is first investigated, then NCSs with stochastictime delay and stochastic packet dropout is modeled as a continuous-time system withtime-varying delays and stochastic parameters. By constructing a more complicatedLyapunov-Krasovskii function and adopting the divided-delay method, a less conservativecriterion for exponential mean-square stability of networked control systems withstochastic time delay and stochastic packet dropouts is presented.Moreover, on the problem of stability analysis of NCSs, most of the existing papersaimed at designing a reasonable controller so as to overcome the negative factors causedby communication network. In this thesis, however, in addition to the negative factorscaused by networked, the deterioration of mechanical equipment is also taken intoconsideration, and a more realistic model is derived. By utilizing the stochastic Lyapunovstability theory, sufficient conditions for the closed-loop system to reach exponentialmean-square stability with prescribed H performance constraint are provided in formsof linear matrix inequalities (LMIs).