This paper proposes a systematic designs methodology for the Takagi-Sugeno (TS) model based fuzzy control systems with guaranteed stability and pre-specified transient performance for the application to a nonlinear magnetic bearing system. More significantly, in the proposed methodology , the control design problems which considers both stability and desired transient performance are reduced to the standard LMI problems . Therefore, solving these LMI constraints directly (not trial and error) leads to a fuzzy state-feedback controller such that the resulting fuzzy control system meets above two objectives. Simulation and experimentation results show that the proposed LMI-based design methodology yields only the maximized stability boundary but also the desired transient responses.

Optical scanning systems use glavanometers to point the laser beam to the desired position on the workpiece. The angular speed of a galvanometer is typically controlled using Proportional+Integral+Derivative(PID) control algorithms. However, natural variations in the dynamics of different galvanometers due to manufacturing, aging, and environmental factors(i.e., process uncertainty) impose a hard limit on the bandwidth of the galvanometer control system. In general, the control bandwidth translates directly into efficiency of the system response. Since the optical scanning system must have rapid response, the higher control bandwidth is required. Auto-tuning PID algorithms have been accepted in this area since they could overcome some of the problems related to process uncertainty. However, when the galvanometer is attached to a larger mechanical system, the combined dynamics often exhibit resonances. It is well understood that PId algorithms may not have the capacity to increase the control bandwidth in the face of such resonances. This paper compares the achieable performance and robustness of a galvanometer control system using a PID controller tuned by the Ziegler-Nichols method and a controller designed by the Quantitative Feedback Theory(QFT) method. The results clearly indicate that-in contrast to PID designs-QFT can deliver a single, fixed controller which will supply high bandwidth design even when the dynamics is uncertain and includes mechanical resonances.

This paper presents a fuzzy H$\infty$ filtering problem for a class of uncertain nonlinear systems with time-varying delayed states and unknown inital state on the basis of Takagi-Sugeno(T-S) fuzzy model. The nonlinear systems are represented by T-S fuzzy models, and the fuzzy control systems utilize the concept of the so-called parallel distributed compensation. Using a single quadraic Lyapunov function, the stability and L2 gain performance from the noise signals to the estimation error are discussed. Sufficient conditions for the existence of fuzzy H$\infty$ filters are given in terms of linear matrix inequalities (LMIs). The filtering gains can also be directly obtained from the solutions of LMIs.

This paper presents an H$\infty$ controller design method for linear systems with time-varying delayed states, inputs, and measurement outputs. Using a Lyapounov unctional , the stability for delay systems is discussed independently of time delays . And a sufficient condition for the existence of H$\infty$ controllers of n-th order is given in terms of three matrix inequalities. Based on the positive-definite solutions of their matrix inequalities, we briefly explain how to construct H$\infty$ construct H$\infty$ controller, which stabilizes time-delay systems independently of delays and guarantees an H$\infty$ norm bound.

One of the important challenges facing control engineers in developing automated machineryis to be able to monitor the machines using remote sensors. Observrs are often used to reconstruct the machine variables of interest. However, conventional observers are unalbe to observe the machine variables when the machine models, upon which the observers are based, have inputs that cannot be measured. Since this is often the case, the authors previsously developed a steady-state optimal state and input observer for time-invariant systems [1], this paper extends that work to time-variant systems. A recursive observer, similar to a Kalman-Bucy filter, is developed . This optimal observer minimizes the trace of the error variance for discrete , linear , time-variant, stochastic systems with unknown inputs.

In this paper, we present robust reliable H$\infty$ controller design methods of continuous and discrete uncertain time delay systems using LMI (linear matrix inequality) technique, respectively. Also the existence conditions of state feedback control are proposed . Using some changes of variables and Schur complements, the obtained sufficient conditions are transformed into an LMI form. The closed loop system by the obtained controller is quadratically stable with H$\infty$ norm bound for all admissible uncertainties, time delay, and all actuator failures occurred within the prespecified set. We show the validity of the proposed method through numerical example.

This paper presents a new delay-dependent robust stabilization condition for uncertain time-delay systems. An algorithm involving convex optimization is proposed to compute a suboptimal upper bound of the delay such that the system can be stabilized by the controller for all admissible uncertainties. It is illustrated by numerical examples that the proposed delay-dependent controller can be less conservative than previous results. It is also shown that the proposed delay-dependent controller can even capture the delay-independent stability of the system, which is not possible with existing delay-dependent results.

The robust generalized H2 filtering problem for a class of discrete time uncertain linear systems satisfying the sum quadratic constraints(SQCs) is considered. The objective of this paper is to develop robust stability condition using SQCs and design a robust generalized Ha filter to take place of the existing robust Kalman filter. The robust generalized H2 filter is designed based on newly derived robust stability condition. The robust generalized Ha filter bounds the energy to peak gain from the energy bounded exogenous disturbances to the estimation errors under the given positive scalar ${\gamma}$. Unlike the robust Lalman filter, it does not require any spectral assumptions about the exogenous disturbances . Therefore the robust generalized H2 filter can be considered as a deterministic formulation of the robust Kalman filter. Moreover, the variance of the estimation error obtained by the proposed filter is lower than that by the existing robust Kalman filter. The robustness of the robust generalized H2 filter against the uncertainty and the exogenous signal is illustrated by a simple numerical example.

This paper describes the communication method of an auto-startup system. The Auto-Startup system is designed to operate a nuclear power plant automatically during the startup operation . In general , the operations during startup in existing plant have only been manually controlled by the operator. The manual operation caused to the operator mistake. The Auto-Startup system consists of the Distributed Control System(DCS) and G2 (Expert System). Also, Functional Test Facility(FTF) provides the plant's real-data for an Auto-Startup system. So, it is necessary to develop the communication method between these systems. We developed two methods ; one is a network and the other is a hardwire line. To communicate between these systems (DCS-G2 and DCS-FTF) , we developed the communication program. In case of DCS-FTF, we used the TCP/IP and VXI. BUt, in case of DCS-G2 , we , what it called , developed the bridge program using the GSI(G2 Standard Interface). We test to check the function of the important parameter, in time, for analysis of the developed communication method. The results are a good performance when we check the communication time of important parameter. We conclude that Auto-startup system could save heat-up time about at least 5 hours and reduced the change of the reactor operation and trip.

It is an interesting area in the field of artifical intelligence to find an analytic model of cooperative structure for multiagent system accomplishing a given task. Usually it is difficult to design controllers for multi-agent systems without a comprehensive knowledge about the system. One of the way to overcome this limitation is to implement an evolutionary approach to design the controllers. This paper introduces the use of a genetic algorithm to discover a fuzzy logic controller with rules that govern emergent agents solving a pursuit problem in a continuous world. Simulation results indicate that, given the complexity of the problem, an evolutionary approach to find the fuzzy logic controller seems to be promising.