서울의 Penicillinase Producing Neisseria gonorrhoeae 발생빈도(1998)

김재홍,김준호,반재용,이정우,황성주,정준규,정성태,강진문,조흔정,홍창의,정혜신,이한승,김이선,이봉길,이종호,선영우,한기덕,윤성필,이성훈,안종성,박석범,문승현,조항래,김형섭,류지호,황재영,박준홍,손상욱 한양대학교 의과대학 2001 한양의대 학술지 Vol.21 No.1

In recent years, gonorrhea has been pandemic and remains one of the most common STDs in the world, especially in developing countries. For the detection of a more effective therapeutic regimen and assessing the prevalence of Penicillinase Producing Neisseria gonorrhoeae(PPNG), we have been trying to study the patients who have visited the Venereal Disease Clinic of Choong-Ku Public Health Center in Seoul since 1980 by menas of the chromogenic cephalosporin method. In 1998, 93 strians of N. genorrhoeae were isolated, among which 60(64.5%) were PPNG. The prevalence of PPNG in Seoul, which had been decreased to 39% in 1996 after a peak of 74.3% in 1993, is increased to 64.5% in 1998.

적응 역 전파 신경회로망 은닉층 노드 수의 최적화에 관한 연구

홍봉화 세명대학교 2000 世明論叢 Vol.8 No.-

The algorithm for improving the speed of convergence and learning rate of back propagation neural networks was proposed by the reference 9) which proposed algorithm updates the learning rate parameter and momentum term for each weight by generated error, adaptively. On the X-OR, 3-parity problem this method show that when compared to the conventional back propagation algorithm, the improved algorithm reduces the number of training iterations about 20%~40%, But, the proposed algorithm didn't to consider about number of node for the hidden layer unit. Therefore, we carry out researches into the proposed algorithm and the number of hidden node which is very important play a rule to the learning. In this paper, Simulation on the X-OR, 3-parity problem this algorithm show that when according to the number of hidden node, proposed algorithm different from the number of training iterations about 0~30% times and adaptive learning parameter inverse proportion for node number of hidden layer.

記憶素子材料로서의 强磁性薄膜에 關한 硏究

洪鳳植 충남대학교 공업기술개발연구소 1975 工業技術開發硏究所論文集 Vol.2 No.2

The influence of pH in electrodeposited solution upon ferromagnetic thin films are studied in order to decide the proper production condition for good practical and stable quality of ferromagnetic thin film which can be applied as memory elements produced by the electrodeposited method. The thin films of 2000Å thickness of Ni, Permalloy, Co, and Fe are made on polycrystalline Cu substrate by the electrodeposited method, and the magnetic properties such as saturation magnetization and coercive force are measured with the magnetic hysteresis loop tracer modified for this work : another test is made on the effect of heat treatment of the thin films, after annealing is applied in the temperature range of 100-400℃ in 2×10^-5 Torr. The values of saturation magnetization the films of Ni, Permalloy, Co, and Fe films deposited in pH less than 6.5 reached to those of bulk and the values of saturation magnetization of the films deposited in pH more than 6.5 decreased to half of those of pH than 6.5. The coercive force of the thin films of Ni, Permalloy, Co, and Fe of pH less than 6.5 is as follows : Ni ; about 70 Oe Permally ; about 12 Oe Co ; about 50 Oe Fe ; about 30 Oe The values of coercive force of pH more than 6.5, excepting the Fe film rose 3 to 4 time higher values than those of pH less than 6.5. On the other hand, the coercive force of Fe, whose pH was more than 6.5 has the tendency to decrease. If the electrodeposited films of pH more than 6.5 are annealed at 350℃ for an hour, the saturation magnetization becomes almost the same value as that of the films of pH less than 6.5 except Fe which shows a slight decrease with increase of temperature. Experimental results show that any saturation magnetization and sudden change of coercive force in either above or below 6.5 of pH occurs when hydrogen elements of the solution insert into the grain boundary or when hydrogen oxide occurs in it. Consequently, according to the experimental results, it may be concluded that in order to have films which can be applied as stable memory elements produced by electrodeposited method, the films are to be annealed for an hour at 350℃ temperature in vacuum, keeping the pH of the electrodeposited solution below 6.5..

전자공학교육과 실험교과내용에 관한 연구

홍봉식,김태균,오승엽,김환우,권오석 충남대학교 공업교육연구소 1983 論文集 Vol.6 No.2(A)

The currenty offered experiment courses are completely reconsidered and reconfigured. The whole subjects are divided into 10 seperate courses, each of which should represent either basic or advanced stages of differing fields of the electronic engineering. It has been aimed that the courses form an elaborate hierachy so that they are minimally redundant but linked to one another in their fringes. Also, it has been pursued that the new courses should not fall obsolete in at last next five years exploring the maximum possibilities of the currently available equipments.

P-형 실리콘에 인(phosphorous)의 선택확산에 관한 연구

홍봉식,구경완 충남대학교 공업교육연구소 1985 論文集 Vol.8 No.1

A phosphorous-diffusion is carried out in open-tube system, using a liquid-pocl₃as a diffusion source. The concentration profiles of phosphorous-diffused layers in P-type silicon wafer have been obtained using a step-etching and sheet resistivity measurement. Junction depth has been obtained by the grooving method and the angle lapping & staining method. About the phosphorous distributions, those show that 1) they do not obey the predicted complementary error-function when the concentration is more than 10^19/cm³, 2) the surface concentration is a constant up to about the half of the depth of resultant junction, 3) the depth of resultant junction is linearly proportaional to the square root of diffusion time, 4) the phosphorous-diffusion coefficient is 1.0 × 10^-13cm²/set at 1,000℃ 5) the critical conditions of gas flows are N₂; 830cc/min, O₂; 45 cc/min, N₂(POCl₃) ; 10 cc/min in pre-deposition 6) the ratios of the logitudinal depth to transverse were increased to the time increament in Selective diffusion.

Photodetector 용 GaAs MESFET의 최적설계에 관한 연구

洪淳錫,朴奉任 홍익대학교 산업기술연구소 2001 産業技術 Vol.11 No.-

In this paper, in order to analyze the optical response for the buried gate type GaAs MESFET photodetector, we derived the total sheet charge density at every channel position. In solving the carrier continuity equation under illumination, we divided the channel region into the depletion and neutral regions. In the depletion region, we considered the drift current only and used the suggested average value for the vertical electric field. While in the neutral region,we neglected the drift current component and suggested some boundary conditions.

콘크리트가 RC 인장부재의 인장거동에 미치는 영향

홍창우,김남윤,윤경구,이봉학 江原大學校 産業技術硏究所 1997 産業技術硏究 Vol.17 No.-

This paper presents evaluation results of the tensile behavior of reinforced high strength concrete. The effects of different sizes of reinforcing bar, ranging from D22 to D29, on the formation of cracks was investigated. Two different strength concretes, 270kg/㎠ and 550kg/㎠, were used in the specimens to investigate the influence if concrete strength on tension stiffening. In the present investigation a method was developed to obtain reliable load-deformation behavior in tension. The experimental results show that (1) high-strength concrete members exhibited larger amounts of tension stiffening than the companion normal-strength concrete members, (2) as the bar diameter increases, the beneficial influence of high-strength concrete on tension stiffening is reduced.

N型 실리콘內의 Boron擴散에 관한 硏究

홍봉식 충남대학교 공업교육연구소 1982 論文集 Vol.5 No.1

The concentration profiles of boron-diffused layers in N-type silicon have been obtained using a steep-etching and sheet resistivity measurement method. About the boron distributions, those show that 1) they do not obey the predicted complementary error-function when the concentration is more than 10^19/㎤, 2) the surface concentration is a constant up to about the half of the depth of resultant junction, 3) the depth of resultant junction is linearly proportional to the sqare root of diffusion time, 4) the boron-diffusion coefficient is 3.21×10 exp^13㎠/sec at 1,000℃.

鮟鱇網 漁具漁法의 改良에 關한 硏究 Ⅴ : 展開帆의 性能

洪聖根,宋之浩,趙鳳坤,朴鍾洙 군산대학교 1981 群山水産專門大學 硏究報告 Vol.15 No.1

We have been developed improved stow net and operated with it in the field. In order to study on the efficiency of otter sail of the improved stow net, we made a model experiment in a circular water tank. The findings are as follows. 1.In proportion to increase water velocity, the height of canvas o.s get lower, the decided fromula is obtained as h(%) = 77.7v-0.087 2.Resistance of otter sail is higher than board type. The hydrodynamic pressure is not perpendicular to the o.s, but inclination backward. 3.The decided formula of developing force, and resistance of otter sail is obtained as table 3. 4.Developing force is 15 to 25% of all resistance of fishing gear, resistance of otter sail is 12-18% and hydrodynamic efficiency is about 1.35. 5.Center of hydrodynamic pressure is 0.38 from the fore edge of otter sail and it may use Helmholtz's formula as it is. 6.The most effective angle of attack is about 26˚, the shearing coefficient at the time is 0.7 and the drag coefficient is 0.5. 7.To increase or decrease the buoyancy and weight of otter sail have not influenced distinctly. 8.Aspect ratio is the most effective in 2.5. 9.The joining position of fore bridles have not influenced on the efficiency of otter sail.

Hopfield 모델의 고속화에 관한 연구

홍봉화,조원경 慶熙大學校 大學院 1994 高凰論集 Vol.14 No.-

A study on implementation of digital neural networks can be the high speed operation. In this paper, the arithmetic processing of Hopfield is able to matrixvctor operation which it is adapt for design to array processor because it has recursive and iterative operation. Also, we have designed to floating point arithmetic of nonlinear function using look up table with very high speed operation. The proposed method, in this paper, is expect to applicative field of real world neural networks because of the realizationable to developed VLSI technique, up to date.