Font Size: a A A

Modeling, Analysis And Design Of Packet-based Networked Control Systems

Posted on:2015-03-15Degree:MasterType:Thesis
Country:ChinaCandidate:D F HeFull Text:PDF
GTID:2268330428497087Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
Networked control systems (NCSs) are systems in which control loops are closed over a communication network which may be share with other applications. As network control systems introducing network into a control system bring many advantages over conventional point-to-point control systems, such as low cost, low weight, low power consumption, easy installation and maintenance and so on, thus gradually applied in industrial processes control, the mobile egent remote control and formation, Aeronautics and astronautics and so on. However, data exchange via network inevitably results in network induced delay, packet loss, limited network access, such as the limited bitrate special circumstances, which results in the decrease of system performance in different ways, even divergence in the system. Therefore these negative factors have to be taken into account in analysis and synthesis of NCSs, and it is known that time-varying packet losses and variable sampling can be equivalently deemed as time-varying transmission intervals. How to compensate transmission delays and intervals in a linear NCSs is a fundamental problem.Based on this, by studying the basis of currently existing theoretical results, the study were made by the following two aspects:First we have concerned with linear packet networked control systems mean square exponential stabilization. Based on a multivariate i.i.d. model of input delays and a Bernoulli model of input missing, the packet-based networked system is formulated as a switching system with multiple subsystems of different input delays. The sufficient fast-switching conditions are directly established for the closed-loop mean-square exponential stability via an appropriate Lyapunov-Krasovskii approach different from average dwell time approach. The resulting controller design method can be obtained by cone complementarity linearization (CCL).Second we have concerned with fuzzy-model-based Hoo design of nonlinear packetized NCSs with time-varying transmission delays and transmission intervals based on a random-delay approach. The real-time distribution of input delays resulting from transmission delays and intervals is taken into account and modeled as a dependent and nonidentically distributed process. A randomly switched Takagi-Sugeno (T-S) fuzzy system with multiple input-delay subsystems is proposed to model the nonlinear packetized NCSs. Based on an improved Lyapunov-Krasovskii method, which appropriately exploits the real-time distribution of input delays in estimating cross-product integral terms as well as the characteristics of T-S fuzzy model, new delay-distribution-dependent sufficient conditions are derived for the deterministic Hoo performance of the over-all systems. The resulting controller design method is equivalent to a nonlinear convex optimization problem with LMI constraints. Numerical examples are presented to substantiate the effectiveness and advantage of our results.Finally, we summarize and prospect to this paper, point out the contribution of this paper and the next research direction.
Keywords/Search Tags:Networked Control Systems (NCSs), Switching Model, Lyapunov-Krasovskii method, Takagi-Sugeno fuzzy model, random-delay approach
PDF Full Text Request
Related items