• Journal of the Institute of Convergence Signal Processing
• /
• v.6 no.1
• /
• pp.45-52
• /
• 2005

Dynamic neural networks have been applied to diverse fields requiring temporal signal processing. This paper describes system identification using the hybrid neural network, composed of locally(LRNN) and globally recurrent neural networks(GRNN) to improve dynamics of multilayered recurrent networks(RNN). The structure of the hybrid nework combines IIR-MLP as LRNN and Elman RNN as GRNN. The hybrid network is evaluated in linear and nonlinear system identification, and compared with Elman RNN and IIR-MLP networks for the relative comparison of its performance. Simulation results show that the hybrid network performs better with respect to the convergence and accuracy, indicating that it can be a more effective network than conventional multilayered recurrent networks in system identification.

• 한국신재생에너지학회:학술대회논문집
• /
• 2005.11a
• /
• pp.647-647
• /
• 2005

• Proceedings of the KIEE Conference
• /
• 1997.07e
• /
• pp.1880-1882
• /
• 1997

구형 다층매질로 모델링된 인체의 두뇌부분에 평면가 입사할 때의 전력흡수 특성을 조사하였다. 사용된 두 주파수인 912MHz 및 2450MHz에서의 조직내의 흡수전력분포를 각각 계산하였다.

• 한국전산유체공학회:학술대회논문집
• /
• 2011.05a
• /
• pp.117-123
• /
• 2011

The low electric power and high efficiency chips are required because of the appearance of smart phones. Also, high-capacity memory chips are needed. e-MMC(embedded Multi-Media Card) for this is defined by JEDEC(Joint Electron Device Engineering Council). The e-MMC memory for research and development is a memory mulit-chip module of 64GB using 16-multilayers of 4GB NAND-flash memory. And it has simplified the chip by using SIP technique. But mulit-chip module generates high heat by higher integration. According to the result of study, whenever semiconductor chip is about 10 $^{\circ}C$ higher than the design temperature it makes the life of the chip shorten more than 50%. Therefore, it is required that we solve the problem of heating value and make the efficiency of e-MMC improved. In this study, geometry of 16-multilayered structure is compared the temperature distribution of four different geometries along the numerical analysis. As a result, it is con finned that a multilayer structure of stair type is more efficient than a multilayer structure of vertical type because a multi-layer structure of stair type is about 9 $^{\circ}C$ lower than a multilayer structure of vertical type.

• Journal of IKEEE
• /
• v.24 no.1
• /
• pp.238-242
• /
• 2020

This paper presents a technique using a multilayered dielectric coating to reduce the radar cross section (RCS) value of an integrated mast mounted in a destroyer. The proposed multilayer structure has the advantage of being easy to fabricate because the dielectric constant is defined so that a general dielectric that does not contain a magnetic component that requires high dielectric constant or is frequently used for blocking electromagnetic wave absorption can be used. After applying the proposed multi-layer dielectric structure to the integrated mast shape, the simulation results show that the RCS reduction performance is 10.9dB at 6GHz, 11.95dB at 12GHz, and 11.63dB at 18GHz compared to the structure without the multilayer structure.

• KIEE International Transactions on Electrophysics and Applications
• /
• v.4C no.4
• /
• pp.149-154
• /
• 2004

This paper presents a silicon probe tip for vertical probe card application. The silicon probe tip was fabricated using MEMS technology such as porous silicon micromachining and deep- RIE (reactive ion etching). The thickness of the silicon epitaxial layers was 5 ${\mu}{\textrm}{m}$ and 7 ${\mu}{\textrm}{m}$, respectively. The width and length were 40 ${\mu}{\textrm}{m}$ and 600 ${\mu}{\textrm}{m}$, respectively. The probe structure was a multilayered structure and was composed of Au/Ni-Cr/Si$_3$N$_4$/n-epi layers. The height of the curled probe tip was measured as a function of the annealing temperature and time. Resistance characteristics of the probe tip were measured using a touchdown test.

• Structural Engineering and Mechanics
• /
• v.75 no.3
• /
• pp.357-367
• /
• 2020

The objective of this article is investigation of dynamic response of thick multilayer functionally graded (FG) beam under generalized dynamic forces. The plane stress problem is exploited to describe the constitutive equation of thick FG beam to get realistic and accurate response. Applied dynamic forces are assumed to be sinusoidal harmonic, sinusoidal pulse or triangle in time domain and point load. Equations of motion of deep FG beam are derived based on the Hamilton principle from kinematic relations and constitutive equations of plane stress problem. The numerical finite element procedure is adopted to discretize the space domain of structure and transform partial differential equations of motion to ordinary differential equations in time domain. Numerical time integration method is used to solve the system of equations in time domain and find the time responses. Numerical parametric studies are performed to illustrate effects of force type, graduation parameter, geometrical and stacking sequence of layers on the time response of deep multilayer FG beams.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.10a
• /
• pp.45-49
• /
• 1997

In this paper, we controlled a Rotary Inverted Pendulum System using Neuro-Fuzzy Controller(NFC). The inverted pendulum system is widely used as a typical example of an unstable nonlinear control system which is difficult to control. Fuzzy theory have been because membership functions and rules of a fuzzy controller are often given by experts or a fuzzy logic control system. This controller is a feedforward multilayered network which integrates the basic elements and functions of a tradtional fuzzy logic controller into a connectionist structure which has distributed learning abilities. Such NFC can be constructed from training examples by learning rule, and the structure can be trained to develop fuzzy logic rules and find optimal input/output membership functions. Using this controller, we presented the results that controlled a Rotary Inverted Pendulum System and the associated algorithms.

• KIEE International Transactions on Electrophysics and Applications
• /
• v.4C no.6
• /
• pp.300-305
• /
• 2004

In this paper, new design equations based on the pole-zero analysis for multi-layered interdigital stripline linear group delay bandpass filter with tap input ports are presented. As a design example, a four-pole group delay filter with center frequency of 2.14GHz, bandwidth of 160MHz, and group delay variation of $\pm$0.1nS for LTCC technology or multilayered PCB technology is designed. In the design process, it is not necessary to simulate the entire structure, as the simulation of half structures is sufficient. Good results can be attained after the optimizing process was performed three times using the proposed equations and a commercial EM simulator.

• Journal of Surface Science and Engineering
• /
• v.28 no.6
• /
• pp.376-385
• /
• 1995

Copper-chromium multilayers with a nominal bilayer thickness of about 400 $\AA$ (200 $\AA$ each) were prepared by dc magnetron sputtering and the evolution of microstructure during heat treatment was investigated by using x-ray diffractometry(XRD), Auger electron spectroscopy(AES) and transmission electron microscopy(TEM). It was observed that an amorphous phase with a thickness of about 40 $\AA$ was formed at the interfaces of the as-deposited Cu/Cr multilayered film using cross-sectional TEM. At elevated temperatures, the Cu(111) reflection showed increasing intensity and decreasing line-width as a result of copper grain growth. The intermixed amorphous phase disappeared after annealing at $250^{\circ}C$ for 1 h and the multilayer structure was stable up to $400^{\circ}C$ for 1 h annealing. At $600^{\circ}C$ annealing, it was observed that the multilayer structure was completely destroyed and copper and chromium phases were fully intermixed.