Research On Optimization Modeling And Control Methodologies Of Networked Control Systems

Networked control systems (NCSs) lie at the intersection of control and com-munication theories. With the development of the communication technology in recent years, the network-based design of control system has emerged as a topic of major interest. Compared with the control system marked by traditional "point to point" direct link, there are several advantages for NCSs due to the introduction of communication network, such as less wire, higher flexibility, easier maintance, and so on. However, the properties of communication network, such as delay, packet dropouts, band-limited channels, etc, can bring many problems for designing the NCSs. These problems may not only deteriorate the performance of control system, but also make the analysis and design of NCSs complex. Therefore, it has become a hot issue how to solve them in NCSs by utilizing effective control techniques. This dissertation makes further research on the modeling, analysis and control of NCSs. The main contents are concluded as follows:1. The modelling and control synthesis problems are studied for a class of NCSs with piecewise constant generalized sampled-data hold function (PCGSHF). A novel NCSs model based on the PCGSHF is constructed, in which both the network-induced delay and the packet dropouts are considered in the trans-mission. A multi-datum point parameter method is proposed for handling the design of both the PCGSHF and the networked feedback controller. By using this method, the design of them is equivalent to determine the bound uncertainty weight of PCGSHF for ensuring the stability of NCSs, which is based on the given finite multi-datum points of PCGSHF. An effective norm bounded method based on the segments of PCGSHF is proposed for handling the nonlinear term which induced by the inter-sampler behavior. The sufficient condition for ensuring the stability of the NCSs can be obtained based on the Lyapunov theory and LMI technique. Both the PCGSHF and the feedback controller are designed.2. The periodic feedback control problem for sampled-data system with input delay is studied. The obtained results are extended to the design of NCSs with different kinds of generalized sampled-data hold function (GSHF). Based on the Fourier series of periodic signal, a novel nonlinear piecewise periodic function (NPPF) is proposed to construct the GSHF. A hybrid nonlinear delay system model is constructed by establishing the relationship between the input delay of system and the frequency of the NPPF. Then, the asymptotic stability criterion can be obtained based on the Lyapunov theory and LMI technique. The sufficient and necessary condition that ensures the GSHF global smooth-ness is proposed. By using this strategy, the jump properties of GSHF is removed, and the reliability of GSHF in practical applications is increased.3. The modelling and control synthesis problem of NCSs with band-limited chan-nels is studied. Firstly, since there occur packet dropouts when the informa-tion is transmitted through the band-limited channels, a novel compression compensation(C-C) method is proposed to ensure the integrity of transmis-sion. Based on this method, a nonlinear switched system model with uncer-tainty is constructed, in which both the inter-sampler behavior and the packet dropout behavior are considered. Furthermore, some more general stability results of the NCSs with band-limited channels are obtained by using the Lya-punov theory and LMI technique, and then the design of both state feedback controller and compensator could be available.4. For periodic feedback control problems of networked control systems (NCSs) with stochastic time-delay, a novel Markovian stochastic time-delay system model is proposed. This model contains a piecewise constant generalized sampled-data hold function (PCGSHF) and a Markov chain which is used describe the existence of network-induced delay and packet dropouts. By us-ing Lyapunov method and the linear matrix inequality (LMI) technique, the sufficient condition for ensuring the closed-loop NCSs stochastically stable is obtained. The design procedures of both the PCGSHF and the state feedback controller are also proposed.5. The stability and controller design problem of stochastic event-driven net-worked control systems (NCSs) is studied. The work process of NCSs are divided into two independent states:one is the successful transmission state under network-induced delay, the other is the packet dropout state which means the the packet can not be received within schedule time. The difference between these two states is if the transmission time exceed the maximum allowable delay. A event-based Markovian chain is proposed the describe jump between every successful transmission of the system, which take both the network-induced delay and packet-dropouts into consideration. The NCSs is modelled as a hybrid nonlinear stochastic switching system. In terms of the nonlinear disturbance caused by the inter-sample behavior, two new norm bounded methods are presented. Based on the Lyapunov theory and LMI technique, sufficient conditions for the mean square stability of the system are obtained and the corresponding control laws can be obtained.Finally, concluding remarks are given. Some unsolved problems and develop-ment direction of NCSs are proposed. Furthermore, the prospects of the further study are discussed.