This paper deals with the displacement estimation of magnetically suspended simple 1 DOF(degree of freedom) system without the displacement sensor. Inherently electro-magnet for control has two natural feedback loops. One is the transfer function which represents the dependance of the amount of the magnetic flux on the gap displace-ments. The other is the transfer function expressing the properties that the back electromotive force is derived from the time derivative of the magnetic flux. Through these two feedback loops, information about the gap length can be represented by the magnetic flux and the coil current. This means that the gap length can be detected from these two states variables of the electromagnet without a displacements sensor(self-sensing). The displacement can be estimated with the magnetic flux subtracted by the coil current. In this paper we use a balance beam in order to deal with the displacement sensorless estimation of the magnetic bearing system. For the stable estimation of the gap displacements by using the method of self-sensing simple PD controller is used. We first show the mathematical model of the balance beam, and then we show the effectiveness of the current and flux feedback for making stable estimation of the gap displacements for the balance beam. Simulation results show the effectiveness of the current and flux feedback for good estimation of the displacement without using displacement sensor.

In this paper, a fuzzy logic controller is designed for position control of an Arage disk. The Arage disk system is an experimental set to exploit Arago's disk phenomenon which is the operation principle of induction motors. Since the Arage disk system operates in stable, maginally stable, and unstable regions, it is suitable as a test system to evaluate efficiency of various control system design methods. It is shown that the fuzzy logic controller shows good responses for multi-operating points of Arage disk system, while the controllers using linearized models are able to control the system on only one operating point.

This paper presents the new adaptive optimal scheme for the nonlinear systems, which is based on the Picard's iterative approximation and fast Walsh transform. It is well known that the Walsh function approach method is very difficult to apply for the analysis and optimal control of nonlinear systems. However, these problems can be easily solved by the improvement of the previous adaptive optimal scheme. The proposed method is easily applicable to the analysis and optimal control of nonlinear systems.

To flexibly evaluate performance and reliability of an electric power system in the aspect of the real-time system which is intrinsically characterized by stringent timing constraints fails catastrophically if its control input is not updated by its digital controller computer within a certain time limit called the hard deadline, we propose fuzzy-random variables and build a discrete event model embedded with fuzzy-random variables. Also, we adapt fuzzy-variables to a path-space approach, which derives the upper and lower bounds of reliability by using a semi-Markov model that explicitly contains the deadline information. Consequently, we propose certain formulas of state automata properly transformed by fuzzy-random variables, and present numerical examples applying the formulas as well.

This paper presents a decentralized compensator design procedure for large scale systems. By using a coordinate transformation method, interconnection terms in t도 subsystem are decoupled and a decentralized observer is designed independently as a form of Luenberger-type for each subsystem. And the decentralized compensator which can select the observer and controller gain independently is designed as a form of observer based controller for each subsystem. The proposed method has a simplicity and conciseness in the aspects of the determination of coordinate transformation matrices and the derivation of the separation principle which is adopted independently for each subsystem.

In this paper, we propose Fuzzy Polynomial Neural Networks(FPNN) based on Polynomial Neural Networks(PNN) and Fuzzy Neural Networks(FNN) for model identification of complex and nonlinear systems. The proposed FPNN is generated from the mutually combined structure of both FNN and PNN. The one and the other are considered as the premise part and consequence part of FPNN structure respectively. As the consequence part of FPNN, PNN is based on Group Method of Data Handling(GMDH) method and its structure is similar to Neural Networks. But the structure of PNN is not fixed like in conventional Neural Networks and self-organizing networks that can be generated. FPNN is available effectively for multi-input variables and high-order polynomial according to the combination of FNN with PNN. Accordingly it is possible to consider the nonlinearity characteristics of process and to get better output performance with superb predictive ability. As the premise part of FPNN, FNN uses both the simplified fuzzy inference as fuzzy inference method and error back-propagation algorithm as learning rule. The parameters such as parameters of membership functions, learning rates and momentum coefficients are adjusted using genetic algorithms. And we use two kinds of FNN structure according to the division method of fuzzy space of input variables. One is basic FNN structure and uses fuzzy input space divided by each separated input variable, the other is modified FNN structure and uses fuzzy input space divided by mutually combined input variables. In order to evaluate the performance of proposed models, we use the nonlinear function and traffic route choice process. The results show that the proposed FPNN can produce the model with higher accuracy and more robustness than any other method presented previously. And also performance index related to the approximation and prediction capabilities of model is evaluated and discussed.

The popular methods to check the identity of individuals include passwords and ID cards. These conventional method for user identification and authentication are not altogether reliable because they can be stolen and forgotten. As an alternative of the existing methods, biometric technology has been paid much attention for the last few decades. In this paper, we propose an efficient system for recognizing the identity of a living person by analyzing iris patterns which have a high level of stability and distinctiveness than other biometric measurements. The proposed system is based on wavelet transform and a competitive neural network with the improved mechanisms. After preprocessing the iris data acquired through a CCD camera, feature vectors are extracted by using Haar wavelet transform. LVQ(Learning Vector Quantization) is exploited to classify these feature vectors. We improve the overall performance of the proposed system by optimizing the size of feature vectors and by introducing an efficient initialization of the weight vectors and a new method for determining the winner in order to increase the recognition accuracy of LVQ. From the experiments, we confirmed that the proposed system has a great potential of being applied to real applications in an efficient and effective way.

We propose an nonlinear demodulation algorithm for undersampled multicomponent AM(Amplitude Modulation), FM(Frequency Modulation) and AM-FM signals. First, we derive respectively undersampling frequency of the AM, FM and AM-FM using undersampling scheme, and separate respectively monocomponent signals from multicomponent signals using periodic algebraic separation algorithm. In this case augmented separation matrix is very regular and sparse, it has a special structure. The proposed demodulation algorithm detects respectively message signals of the IA(Instantaneous Amplitude) and IF(Instantaneous Frequency) from descrete monocomponent AM, FM and AM-FM signals with an undersampling frequency to be controllable. Verifying the RMS(Root Mean Squares) errors of the detected signals, we show that the performance is excellent.