• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.9
• /
• pp.1793-1801
• /
• 1994

In this paper, a new method for predicting the 1 dB gain compression point of cascaded N amplifiers is proposed. With the proposed method, the transfer function of each amplifier is derived from scalar data available from the manufacturers data sheet and all transfer functions are producted with scalar in order to also derive the overall transfer function of the subsystem under the assumption that the input and output port of each amplifier are matched. Therefore, the 1 dB gain compression point of the subsystem can be predicted or estimated, reversely, utilizing the overall transfer function obtained with the proposed method. The proposed method can be used irrespective of the number of scalar data but, in this paper, it is analyzed only with two scalar data (linear power gain and 1 dB gain compression point) and three scalar data(linear power gain, 1 dB and 0.5 dB gain compression points). With two sample amplifiers operated in Ku-band, the predicted results by the proposed and previous method, respectively, and the experimental results are together presented in order to confirm its utility.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.7 no.5
• /
• pp.34-45
• /
• 1997

In this paper, we present hardware implemcntation of self-organizing feature map (SOFM) neural networkwith constant adaptation gain and binary reinforcement function on FPGA. Whereas a tnme-varyingadaptation gain is used in the conventional SOFM, the proposed SOFM has a time-invariant adaptationgain and adds a binary reinforcement function in order to compensate for the lowered abilityof SOFM due to the constant adaptation gain. Since the proposed algorithm has no multiplication operation.it is much easier to implement than the original SOFM. Since a unit neuron is composed of 1adde $r_tracter and 2 adders, its structure is simple, and thus the number of neurons fabricated onFPGA is expected to he large. In addition, a few control signal: ;:rp sufficient for controlling !he neurons.Experimental results show that each componeni ot thi inipiemented neural network operates correctlyand the whole system also works well.stem also works well.

• Journal of IKEEE
• /
• v.6 no.2 s.11
• /
• pp.128-135
• /
• 2002

We use the saturation nonlinear feedback element to generate self-oscillation in order to find an ultimate gain and period of linear plant. The use of saturation nonlinear feedback element can improve accuracy of an ultimate gain and period of unknown linear plant. An ultimate gain and period of linear plant can be used to tune a PID controller parameters. It is possible that an asymmetric oscillation can be occurred under the special circumstances such as with static load disturbance. We analyze an asymmetric self-oscillation. As the results of an analysis, we propose a method to find an ultimate gain and period of linear Plant under the asymmetric self-oscillation.

• Journal of Energy Engineering
• /
• v.5 no.1
• /
• pp.28-33
• /
• 1996

This study was carried out find out the appropriate tuning method of PID controller for a package type gas engine cogeneration system in terms of stabilizing the engine coolant temperature and system heat balance. In order to acquire the proper parameters of the controller, a system transfer function was set as a first order plus dead time model and thereafter model parameters were determined by using several tuning methods. And, with determined values of parameters and the system transfer functions, optimal turning method was selected by simulating the process using MATLAB. From the experimental results, it was found that obtained PID gains made the system stable in various operating conditions.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.38 no.6
• /
• pp.26-37
• /
• 2001

In order to improve the several problems of the general backpropagation, we propose a method using a fuzzy logic system for automatic tuning of the activation function gain in the backpropagation. First, we researched that the changing of the gain of sigmoid function is equivalent to changing the learning rate, the weights, and the biases. The inputs of the fuzzy logic system were the sensitivity of error respect to the last layer and the mean sensitivity of error respect to the hidden layer, and the output was the gain of the sigmoid function. In order to verify the effectiveness of the proposed method, we performed simulations on the parity problem, function approximation, and pattern recognition. The results show that the proposed method has considerably improved the performance compared to the general backpropagation.

• The Journal of the Acoustical Society of Korea
• /
• v.4 no.3
• /
• pp.8-15
• /
• 1985

본 논문에서는 코클리어 기저막 운동에 관한 전기적 모델링을 래더회로를 이용하여 구현하였다. Budak 의 이론에 근거한 이득이 1보다 큰 RC 수동회로를 이용하여 귀에서의 Place Principle을 만족시 키는 특성을 갖는 모델을 제시하였으며 이는 간단한 회로를 첨가시킬시에 이득이 1보다 큰 주파수를 검 지해낼 수도 있다. 컴퓨터 Simulation을 이용한 전달함수 분석에서는 특이한 형태 의 전압전달함수를 얻 을 수 있었으며, 연구결과, 본 모델은 코클리어 기저막운동을 표현하기에 적합하다는 결론을 얻을 수 있 었다. 본 모델은 임상의학이나 그 외의 공학분야등에 이용이 가능하리라고 본다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04b
• /
• pp.310-312
• /
• 2002

반복적 죄수의 딜레마게임은 복잡한 사회현상들을 모델링 하기 위하여 주로 사용되는 방법이다. 본 논문에서는 집단을 협동으로 진화시키는데 있어서 전략적 연합을 이용하는 방법과 전략적 연합에 속해 있는 전략들이 연합의 의사를 결정할 때 어떠한 방법을 선택할 경우 더 빠르고 안정적으로 진화하는가에 대한 실험 결과를 제시한다. 의사결정방법으로는 신경망으로 결합에 주로 사용되는 투표방법 Borda 함수, Condorect 함수, 평균방법 그리고 최고이득 함수방법 등을 사용하였다. 실험결과 최고이득 함수 방법이 가장 좋은 결과를 보였으며 결합 방법에 따라 다소 상이한 결과를 나타내었다.

• Journal of IKEEE
• /
• v.2 no.2 s.3
• /
• pp.263-269
• /
• 1998

We use the slope bounded saturation nonlinear feedback element instead of relay to find ultimate gain and period of linear plant. Saturation nonlinear element reduces the high harmonics of plant output. The reduction of high harmonics improve the accuracy of describing function method used to find ultimate gain and period. We give a simple procedure to find ultimate gain and period with saturation nonlinear element. A PID controller design method with known time delay element is also given, which is very useful when oscillation is not occurred with nonlinear element.

• Korean Journal of Optics and Photonics
• /
• v.14 no.3
• /
• pp.279-285
• /
• 2003

Wide- and flat-gain laser diodes were designed and fabricated from asymmetric multiple quantum well (AMQW) structures which consist of three compressively strained InGaAsP wells of different thicknesses. For a 400 ${\mu}{\textrm}{m}$-long lasers with as-cleaved facets, -1 ㏈ and -3 ㏈ gain bandwidth were 45 nm and 80 nm, respectively. For this AMQW structure, calculated gain spectra with various line broadening functions were compared with experimental results. We confirmed the calculated gain spectra using an asymmetric line broadening function were in good agreement with the measured data.

• Journal of KIISE:Software and Applications
• /
• v.27 no.3
• /
• pp.257-265
• /
• 2000

There is much selective confliction in nature where selfish and rational individuals exists. Iterated Prisoner's Dilemma (IPD) game deals with this problem, and has been used to study on the evolution of cooperation in social, economic and biological systems. So far, there has been much work about the relationship of the number of players and cooperation, strategy learning as a machine learning and the effect of payoff functions to cooperation. In this paper, We attempt to investigate the cooperative coalition size according to payoff functions, and observe the relationship of localization and the evolution of cooperation in NIPD (N-player IPD) game. Experimental results indicate that cooperative coalition size increases as the gradient of the payoff function for cooperation becomes steeper than that of defector's payoff function, or as the minimum coalition size gets smaller, Moreover, the smaller the neighborhood of interaction is, the higher the cooperative coalition emerges through the evolution of population.