Takashi Yahagi

FM/PWM control scheme in class DE inverter

This paper presents a new control scheme for a Class DE inverter, that is, frequency modulation/pulsewidth modulation (FM/PWM) control. Further, the FM/PWM controlled Class DE inverter is analyzed and we clarify performance characteristics. Since the FM/PWM controlled inverter has two control parameters, namely, the switching frequency and the switch-on duty ratio, it has one more degree of freedom for the control than the inverter with the conventional control scheme. The increased degree of freedom is used to minimize the switching losses. Therefore, it is possible to control the output power with high power-conversion efficiency for wide-range control. Carrying out the circuit experiments, we confirm that the experimental results agree well with the theoretical predictions quantitatively. For example, the proposed controlled inverter can control the output voltage from 56% to 191% of the optimum one, which is designed for 1.8 W at 1.0 MHz, with maintaining over 90% power-conversion efficiency.

A Method of Face Recognition Based on Fuzzy c-Means Clustering and Associated Sub-NNs

The face is a complex multidimensional visual model and developing a computational model for face recognition is difficult. In this paper, we present a method for face recognition based on parallel neural networks. Neural networks (NNs) have been widely used in various fields. However, the computing efficiency decreases rapidly if the scale of the NN increases. In this paper, a new method of face recognition based on fuzzy clustering and parallel NNs is proposed. The face patterns are divided into several small-scale neural networks based on fuzzy clustering and they are combined to obtain the recognition result. In particular, the proposed method achieved a 98.75% recognition accuracy for 240 patterns of 20 registrants and a 99.58% rejection rate for 240 patterns of 20 nonregistrants. Experimental results show that the performance of our new face-recognition method is better than those of the backpropagation NN (BPNN) system, the hard c-means (HCM) and parallel NNs system, and the pattern-matching system

Phase control for resonant DC-DC converter with class-DE inverter and class-E rectifier

In this paper, a phase control scheme for Class-DE-E dc-dc converter is proposed and its performance is clarified. The proposed circuit is composed of phase-controlled Class-DE inverter and Class-E rectifier. The proposed circuit achieves the fixed frequency control without frequency harmonics lower than the switching frequency. Moreover, it is possible to achieve the continuous control in a wide range of the line and load variations. The output voltage decreases in proportion to the increase of the phase shift. The proposed converter keeps the advantages of Class-DE-E dc-dc converter, namely, a high power conversion efficiency under a high-frequency operation and low switch-voltage stress. Especially, high power conversion efficiency can be kept for narrow range control. We present numerical calculations for the design and the numerical analyses to clarify the characteristics of the proposed control. By carrying out circuit experiments, we show a quantitative similarity between the numerical predictions and the experimental results. In our experiments, the measured efficiency is over 84% with 2.5 W output power for 1.0-MHz operating frequency at the nominal operation. Moreover, the output voltage is regulated from 100% to 39%, keeping over 57% power conversion efficiency by using the proposed control scheme.

An image analyzing system using an artificial neural network for evaluating the parenchymal echo pattern of cirrhotic liver and chronic hepatitis

To objectively evaluate the parenchymal echo pattern of cirrhotic liver and chronic hepatitis, the authors applied an image analyzing system (IAS) using a neural network. Autopsy specimens in a water tank (n=13) were used to examine the relationship between the diameter of the regenerative nodule and the coarse score (CS) calculated by IAS. CS was significantly correlated with the diameter of the regenerative nodule (p<0.0001, r=0.966). CS is considered to be useful for evaluating the coarseness of the parenchymal echo pattern.

Face recognition using neural networks with multiple combinations of categories

Neural networks trained via backpropagation are now widely applied in a pattern recognition method. However, since it becomes much more difficult for a network to accomplish its task when the number of categories increases, research on multiple network combination is active. Novel learning and recognition processes are proposed here. It is shown that by combining small-scale neural networks, the proposed method allows exploitation of the potential capabilities of the networks. In the learning process, multiple networks are trained with patterns organized in overlapping groups. During the recognition process, response is obtained by making the networks compete with each other. In experiments involving recognition of individuals from various facial images and different expressions, a recognition rate of higher than 96 percent was obtained for 20 individuals and 131 images. Furthermore, results of simulations in which noise was added confirmed that the proposed method is robust with respect to pattern changes.

A method of face recognition based on fuzzy clustering and parallel neural networks

This paper presents a method for face recognition based on fuzzy clustering and parallel neural networks. Neural networks have been widely used in various fields. However, the computing efficiency decreases rapidly if the scale of the neural network (NN) increases. In this paper, a new method of face recognition based on the neuron-fuzzy system is proposed. In particular, the face patterns are divided into several small-scale parallel neural networks based on fuzzy clustering, and they are combined to obtain the recognition result. The proposed method achieves 98.71% recognition accuracy using 310 frontal face images and 98.06% recognition accuracy using 310 rotated face images corresponding to 31 individuals.

Design of generalized class E2 dc/dc converter†

SUMMARY This paper presents a novel design procedure for class E 2 dc=dc converter. The design procedure requires only circuit equations and design specications. When the circuit equations are got, the other procedures for the computation of the design values are carried out with aid of computer. Therefore, we can design class E 2 dc=dc converters with any conditions by using the proposed design procedure. Moreover, we give the design and the performance curves of class E 2 dc=dc converter and discuss about them. By carrying out the circuit experiments, we show the validity of the proposed design procedure. Copyright ? 2003 John Wiley & Sons, Ltd.

A Method of Simple Adaptive Control for Nonlinear Systems Using Neural Networks

This paper presents a method of simple adaptive control (SAC) using neural networks for a class of nonlinear systems with bounded-input bounded-output (BIBO) and bounded nonlinearity. The control input is given by the sum of the output of the simple adaptive controller and the output of the neural network. The neural network is used to compensate for the nonlinearity of the plant dynamics that is not taken into consideration in the usual SAC. The role of the neural network is to construct a linearized model by minimizing the output error caused by nonlinearities in the control systems. Furthermore, convergence and stability analysis of the proposed method is performed. Finally, the effectiveness of the proposed method is confirmed through computer simulation.

A new transformed input-domain ANFIS for highly nonlinear system modeling and prediction

In two or more-dimensional systems where the components of the sample data are strongly correlated, it is not proper to divide the input space into several subspaces without considering the correlation. In this paper, we propose the usage of the method of principal component in order to uncorrelate and remove any redundancy from the input space or the adaptive-neuro fuzzy inference systems (ANFIS). This leads to an effective partition of the input space to the fuzzy model and significantly reduces the modeling error. A computer simulation for three frequently used benchmark problems shows that ANFIS with the uncorrelation process performs better than the original ANFIS.

Investigation of Class E Amplifier with Nonlinear Capacitance for Any Output Q and Finite DC-Feed Inductance

This paper investigates the design curves of class E amplifier with nonlinear capacitance for any output Q and finite dc-feed inductance. The important results are; 1) the capacitance nonlinearity strongly affects the design parameters for low Q, 2) the value of dc-feed inductance is hardly affected by the capacitance nonlinearity, and 3) the input voltage is an important parameter to design class E amplifier with nonlinear capacitance. By carrying out PSpice simulations, we show that the simulated results agree with the desired ones quantitatively. It is expected that the design curves in this paper are useful guidelines for the design of class E amplifier with nonlinear capacitance.