Filtering Methods To Quantify The Robust Network Control System

In recent years, we have an in-depth study for Networked Control Systems with the development of network technology and NCS has become one of the major problems in the control filed. Network Control Systems use quantizers to quantize signal because of existing network, and filtering problem is as important as control issue in control systems, so robust quantized filtering for network control system has practical significance. In this paper, adding logarithmic quantization and a class of dynamic quantization to Networked Control Systems, and also consider random packet loss, time-varying delay, sensor saturation, media access constraints and other issues. The concrete work in this paper as follows:Firstly, the robust H2/H∞quantized filtering problem for a class of networked control systems with packet loss is studied. The packet loss process is modeled by a Bernoulli distribution white sequence with a known probability and logarithmic quantization is used. Sufficient conditions are derived for the filtering error system to be asymptotically stable, ensure a prescribed H∞performance and make the upper bound H2norm minimization based on Lyapunov stability theory. Finally, obtaining the parameters of filter by linear matrix inequality (LMI) technology.Secondly, the robust H∞finite-horizon quantized filtering problem for a class of network time-varying system with sensor saturation and packet loss is studied. A dynamic quantizer, which scaling the quantization error and quantization range by add a scaling factor, is used in this system. Sufficient conditions are derived for the filtering error system to be finite-horizon random bounded, ensure a prescribed H∞performance and the dynamic quantizer designed method based on Lyapunov stability theory. Finally, obtaining the parameters of filter by linear matrix inequality (LMI) technology.Thirdly, the robust H∞quantized filtering problem for a class of networked control system with communication constraints and random delay is studied. Where only limited communication network nodes can be passed in a sampling period because of communication constraints and the time-varying network delay is bounded. Firstly, the random transfer process is described by a Markov process. Using a Markov state transition matrix to determine which nodes can get transmission; then use TOD scheduling policy to determine which nodes can get transmission. Sufficient conditions are derived for the filtering error system to be asymptotically stable, ensure a given H∞performance and the dynamic quantizer designed method based on switching systems theory for the above two cases respectively, Lyapunov-Krasovskii functional approach. Finally, obtaining the parameters of filter by linear matrix inequality (LMI) technology.Finally, for the proposed methods, using Truetime simulation toolbox and Matlab LMI toolbox to simulation.