제왕절개수술시 시행된 마취의 임상적 고찰

정광우,정규섭,최상대 대한마취과학회 1976 Korean Journal of Anesthesiology Vol.9 No.1

The following are summary of the clinical observations on 184 pregnant women underwent cesarean section under general and spinal anesthesia. 1. Ages of pregnant women ranged from 18 to 45 years The second and third decades. comprised 94% of all cases. 2. 83% of the cases were emergency and others were elective operations. 3. Surgical indieations revealed dystocia in 43%, hemorrhage in 35%, and others in 22%. 4. 92 cases were performed by general anesthesia, 88 cases by spinal and the remains by local. 5. Apgar scores checked one minute after delivery were as follows. That was Poor categorY (0 to 4): 42%% Fair category (5 to 7): 19% Good category (8 to 10): 39% Resuscitation of Apgar score 0 babies were successful in 8 cases.

강우 초기 유출수에 의한 오염물질 제어를 위한 여과장치 개발 및 효과 분석

정광우,박종태,서동일 대한환경공학회 2016 대한환경공학회 학술발표논문집 Vol.2016 No.11

인공 신경 회로망의 학습기능을 이용한 한국어 단모음 인식에 관한 연구 ( A Study on the Recognition of Korean Single-Vowel Using Learning Property of the Artifical Neural Network )

정광우,홍광석,박병철 대한전자공학회 1990 대한전자공학회 학술대회 Vol.1990 No.7

다중 출력 MLP신경망을 이용한 한국어 음성인식에 관한 연구

鄭光宇,朴炳哲 成均館大學校 科學技術硏究所 1995 論文集 Vol.46 No.1

In this paper, to improve the performance of the speaker-independent speech recognition system, a new reural network model and learning algorithm were proposed. Recognition algorithm for performance improvement of the recognition system had two kinds. Several experimental results were given to prove the performance of proposed algotithm on the isolated digit recognition of the speaker- independent. First, the algorithm for extraction of the feature vector was proposed. it imported, to stabilize the sequence of the feature vector to input speech, the 'Inertia Term' between neighbor frames. If, caused the reducing fluctuation components produced on the sequence of the feature vector of speech signal and tine suppressing superposition of probability distribution between feature vectors. As a result, the ability of phoneme discrimination was improved on the speaker independent speech recognition system or 3.9%. Second, the Multi-Output MLP(MO-MLP) neural network structure was proposed. It had, for reflecting the variation of the feature of the phoneme which was produced on the variable phonological environment, the structure which was assigned several output nodes to the same phoneme. So, the convergence characteristic of the neural network could be improved by the proposed structure for complex learning data and this structure could improve the ability of recognition system by generating the multi-template pattern on a neural network for 8.4%.

대차프레임 기준하중이력 분석방법에 관한 연구

정광우 한국철도대학 2008 한국철도대학논문집 Vol.23 No.-

As a consequence of the standardization process developing in Europe, on April 2005 the new European standard EN 13749 was issued by the European standardization body CEN. The norm EN 13749 standardizes and develops the requirements already present in UIC leaflets for test verifications and define all technical requirements for the acceptance process in order to achieve a complete satisfactory design of the bogie. The aim of the norm is to define the complete design process of new railway bogies. It includes design procedures, assessment methods, verification and manufacturing quality requirements. In this study, fatigue analysis of the bogie frame is investigated comparing different approaches between conventional methodology and simulation results based on the VPD(Virtual Product Development).

철도차량운전면허제도 도입에 따른 모의운전연습기의 활용방안에 관한 연구

鄭光宇 한국철도대학 2005 論文集 Vol.20 No.-

Korea National Railroad College developed a new generation education/training railway simulator. A driver's console is given to each student in a class of 6 students. The driver's console is a small-sized simulator for railway crew training use; monitoring and controlling are designed be simply conducted by one instructor only. In addition to the typical Cab equipment, simulated visual equipment through real time computer graphics, simulated sound generator and other simulating functions in each driver console, a structure, driving laws and regulations is also built-in. The instructor console is connected with each driver console through a local area network(LAN). Driving skill and result of each student are automatically diagnosed and statistically recorded. Thus, the student is able acquire driving qualification within a realistic and interesting driving environment. Additionally the instructor can grasp the learning level of students in real-time, with less burden on the instructor.

국내외 기관사 훈련 프로그램의 현황 및 문제점

鄭光宇 철도전문대학교 2001 論文集 Vol.16 No.-

The use of simulators for driver training has led to a new approach to training. It is no longer the instructor who possesses the knowledge as such : he relies on new pedagogical experiences - the concept of discovery in particular - and on this type of tool to encourage trainees to find their feet and assimilate the procedures that are to be applied. Driving simulators therefore make a contribution of the utmost importance to the different phases of driver training, including the acquisition of knowledge and knowhow, an understanding of operating systems and the validation of competencies. However, this does require considerable discipline in the use of these tools and means that instructors must be professionally skilled in both technical and pedagogical spheres. Drawback of traditional training Traditioal training consistes of : - oral messages (lecture-style lessions), - written messages (trade documents with instructions), - support and guidance under the supervision of an instructor. Drawback Acquisition process is almost exclusively mental Instructor assistance only allows the trainee to experience standard situation, which account for a very small proportion of the total instruction. Acquisition is therefore virtually never gestural(almost as if someone had to learn to swim from a book). Acquisition process is not interactive The trainee is not called upon to react (unless questioned by the instructor) and most therefore use his entire concentration to assimilate a purely mental task. Acquisition process is not personalised The trainee is one of a group of learners and the of learning is imposed by the instructor. Trainee has no means of self-appraisal Other than the written tests and questions posed orally by the instructor, the trainee has no means of evaluation his progress.

RLS알고리즘을 이용한 Stop and Go 블라인드 등화기에 관한 연구

정광우 철도전문대학교 1998 論文集 Vol.14 No.-

While the stop and go blind equalizer, in the blind start-up mode, converges more stably and faster than other blind equalizers do, it can not be expected to have satisfactory tracking performance in the application to time-varing fading channel because of the decrease of the opportunities to adjust the tap weights which stems from the algorithm itself. In this study, to make the algorithm suitable to the time-varing channel, we replace the LMS algorithm, which used in the conventional stop and go algorithm, with the RLS algorithm and compare the convergence characteristics and steady state tracking performance of the proposed algorithm with those of the conventional algorithm. According to the simulation results for time-invariant and time-varing channel model, it can be seen that in the blind start-up mode the proposed blind equalizer converges faster than the conventional blind equalizer does and In the steady state it has 0.5dB less MSE than those of the conventional algorithm, thus we can conclude that the performance of the proposed algorithm is superior to that of the conventional blind equalization algorithm using the LMS algorithm.

척수마취후 발생한 혼수환자의 치험예

정광우,박정호,정규섭 대한마취과학회 1976 Korean Journal of Anesthesiology Vol.9 No.1

Authors experienced three cases of comatous patient, accompanied with heada-che and convul-sion follwing spinal anesthesia for hysterectomy, reduction of intuegussuption ,and appendectomy in the private clinics. Three cases suspecting chemical or asceptic meningitis were treated under the proper fluid and symptomatic therapy with oxygen inahan and maintenance of wairway. One case which had cardiopulmonary arrest at the private clinic, died after two days' admision. Two cases recovered within 4 and 7 days without any specific neurological complications .

유전 알고리즘을 이용한 한국어 음성인식에 관한 연구

鄭光宇 철도전문대학교 1997 論文集 Vol.13 No.1

This paper proposes a new speech segmentation method using GA(genetic algorithm). The solution in the general GA can be moving because of stochastic reinsertion, and suffer from the premature convergence problem owing to deficiency of individuals before finding the optimal solution. To cope with these problems and to reduce processing time, we propose the new GBR algorithm and the technique that resolves the premature convergence problem. GBR selects the individual in the child pool that has the fitness value superior to that of the individual in the parents pool. We resolve the premature convergence problem with producing the mutation in the parents population, and propose the new method that removes the small regions in the segmented results. The experimental results show that the proposed segmentation algorithm gives better performance than the ACA ones in Gaussian noise environments.