고농도 암모니아를 함유한 전자 폐수의 암모니아 탈기 실증 연구

손재현,김영희,Jae Hyun Son,Younghee Kim 한국청정기술학회 2023 청정기술 Vol.29 No.4

첨단 전자 제품 산업의 비약적인 발전은 환경적 측면에서 고농도 암모니아 폐수의 증가를 초래했다. 고농도 암모니아 폐수를 안정적으로 처리하기 위해 다양한 방법의 기술이 시도되고 있으나, 지금까지 성공적인 기술이 개발되어 적용되지는 못하고 있다. 본 연구에서는 첨단 전자산업에서 발생하는 고농도 암모니아 함유 폐수에 대하여 밀폐(closed) 순환형 대향류 충전탑 형식의 실증설비를 이용하여 온도, 공기부하율 그리고 폐수부하율 변화에 따른 암모니아성 질소의 제거효율과 제거 특성을 평가하였다. 폐수량 20.8 m³ h^-1, 공기량 18,000 Nm³ h^-1의 운전 조건에서 온도를 45, 50, 55 그리고 60℃로 변경하여 운전한 결과, 암모니아성 질소(NH₃-N)의 제거율은 각각 87.5, 93.4, 96.8 및 98.7%로 온도가 제거율에 미치는 가장 큰 영향 인자임을 알 수 있었다. 공기부하율을 증가시키면 제거율도 증가하나, 흡수탑의 액적(droplet)이 탈기탑으로 유입되어 제거율 증가는 크지 않았다. 폐수 부하율이 변경되어도 제거율은 크게 변하지 않았는데, 이는 제거율에 영향이 없는 것이 아니라, 상대적으로 높은 공기부하율에 기인한 것으로 판단된다. 실증연구를 통해 암모니아 탈기법은 첨단 전자산업에서 발생하는 고농도 암모니아 폐수를 안정적으로 처리할 수 있는 적정한 공법임을 확인할 수 있었다. The rapid advancement of the high-tech electronics industry has led to a significant increase in high-concentration ammonia wastewater. Various methods have been attempted to reliably treat wastewater containing high concentrations of ammonia, but no successful technology has yet been developed and applied. In this study, the removal efficiency and characteristics of ammonia nitrogen was evaluated according to changes in temperature, air loading rate, and liquid loading rate using a closed circulation countercurrent packed tower type demonstration facility for wastewater containing high concentrations of ammonia generated in the high-tech electronics industry. The temperature was varied while maintaining operating conditions of a wastewater flowrate of 20.8 m³ h^-1 and an air flow rate of 18,000 Nm³ h^-1. The results showed that at temperatures of 45,50,55, and 60℃, the removal efficiencies of ammonia nitrogen (NH₃-N) were 87.5%, 93.4%, 96.8%, and 98.7%, respectively. It was observed that temperature had the most significant impact on the removal efficiency of NH₃-N under these conditions. As the air loading rate increases, the removal rate also increases, but the increase in removal efficiency is not significant because droplets from the absorption tower flow into the stripping tower. Even if the liquid loading rate was changed by ±30%, the removal rate did not change significantly. This does not mean that the removal rate was unaffected, but was believed to be due to the relatively high air load rate. Through demonstration research, it was confirmed that ammonia stripping is a reliable technology that can stably treat high-concentration ammonia wastewater generated in the high-tech electronics industry.

기준모델 추종기능을 이용한 퍼지제어기의 자기동조

孫宰顯(Jae-Hyun Son),南文鉉(Moon-Hyon Nam) 대한전기학회 1995 전기학회논문지 Vol.44 No.3

인공팔 제어를 위한 근전신호의 신경회로망을 이용한 기능분류

孫宰顯(Jae-Hyun Son),洪性宇(Sung-Woo Hong),南文鉉(Moon-Hyon Nam) 대한전기학회 1994 전기학회논문지 Vol.43 No.6

모델 기준 적응 퍼지 제어기를 이용한 DC 전동기 제어

손재현,김제홍,Son, Jae-Hyun,Kim, Je-Hong 대한전자공학회 1999 電子工學會論文誌-T Vol.t36 No.4

본 논문에서는 FLC의 적응능력에 대한 단점과 제어규칙 도출의 어려움을 극복하기 위하여 모델기준 적응 퍼지제어기(MRAFC)를 제안하였다. MRAFC는 단순퍼지제어기와 플랜트로 구성된 내부 피드백 루프와 단순 퍼지제어기의 제어규칙을 동조시키는 외부 루프로 구성된다. 기준모델은 기준입력에 대해 전체 제어시스템에 요구되는 제어성능을 특성화하고 수량화하는 퍼지제어기 설계기준으로 사용되었다. 그리고 적응 매카니즘은 FLC 제어규칙을 동조하는 역할을 수행한다. 제안한 알고리즘의 성능은 DC 서보 전동기에 대한 실험에 의해 검증되었다. In this paper, model reference adaptive fuzzy controller (MRAFC) was proposed in order to overcome the difficulty of extracting rules and defects of the adaptation performance in the FLC. MRAFC comprised inner feedback loop consisting of the FLC and plant, and outer loop consisting of an adaptation mechanism which was designed for tuning a control rule of the FLC. A reference-model was used for design criteria of a fuzzy controller which characterizes and quantizes the control performance required in the overall control system. Tuning control rules of FLC is performed by the adaptation mechanism. The performance of proposed algorithm was verified through experiment for the DC servo motor.

폐쇄배전반내 모선의 절연이격거리

손재현(Jae-Hyun Son),김선구(Sun-Ku Kim) 대한전기학회 1996 대한전기학회 학술대회 논문집 Vol.1996 No.7

慶州方言と東京方言の外來語のアクセント

손재현 ( Jae Hyun Son ) 한국일어교육학회 2014 일본어교육연구 Vol.0 No.28

日本語と韓國語には方言によって、意味の區別に役立つアクセントが存在する. 本稿では、アクセント方言の中でも、最も多くの硏究が行われている日本語の東京方言と、東京方言と同樣の多型アクセント體系をもつ韓國語慶州方言の外來語アクセントについて比較對照する. 從來、韓日言語の對照硏究は數多く行われているが、その多くが文法論や語彙論であり、音韻論、特にアクセントの分野では、韓國語のアクセント硏究が日本語のそれほど旺盛に行われていないため、對照できる十分なデ一タがなかった. 本稿では、日本語と韓國語のように高低アクセントをもつ言語の類型的硏究、および對照言語學的考察として、日本語東京方言と韓國語慶州方言のアクセント體系を提示しうえで、兩言語の外來語のアクセント型の特徵を中心に考察を深めた. This paper compares the loanwords accent of the Tokyo dialect of Japanese and that of the Gyeongju dialect in Korean. In the present paper, evidence for this interpretation will be given, along with the structure of the Gyeongju dialect loanwords accent and a comparison with it and that of Tokyo. Looking the Tokyo dialect firstly, it has two oppositions for one-mora words, three oppositions for two-mora words, and so on. It can be represented as Pn=n+1, where ‘Pn’ represent the number of Prosodemes for the accentual unit ‘n’. In the Tokyo accent type, there are n-kernel positions, plus a kernel less type (indicated with an = sign; the ⓞ sign indicates zero-accent), for a total of n+1. The main accent of loanwords of Tokyo dialect are -③ pattern. The present paper examines on the basis of the data collected by the author. The conclusion by Kenstowicz & Sohn (2001) (hereafter K & S(2001)) that the accent of loanwords in the Gyeongju dialect is determined by their syllabic structure. As a result, it is shown that the highest average of the application of the K & S(2001)`s principle is about 70% as regards three or four syllable words. This average is only achieved when the following two conditions regarding the original word of borrowing are taken into account. Those conditions are: 1) if the final consonant cluster of the original word are resolved by two inserted vowels、those vowels are without accent; 2) the heavy syllables caused by liquid gemination are not counted as heavy. Furthermore、as regardstwo or five syllable words、the principle holds only for about 60% of cases. There is no correlation between the syllabic structure and accent as regards monosyllabic words、since the accent remains the same even if the syllabic structure is altered by the addition of a particle. From the above discussion the following points in K &S (2001) are criticized: The pattern "...HLL (antepenultimate)"、which、according to K & S (2001)、is limited to the case when the final consonant cluster of the original is resolved by two inserted vowels、occurs elsewhere too. Their explanation is thus insufficient. They limit their discussion of the accent patterns of the dialect concerned to three types only、namely、"HH... (doubled)"、"...LH (final)" and "...HL (penultimate)"、while other types are simply ignored. Thus the present paper critically examines the applicability of the two conditions on the segmental structure of the original word、the prerequisites of the discussion of K & S (2001)、as well as the level of knowledge of the original words in the acquisition of loanwords by the dialect speakers. Lastly, in this paper, including loanwords accent pattern of the Tokyo and Gyeongju dialects, will be explained in detail on the tonal patterns and the internal mechanisms.

이동 무선 센서 네트워크에서 에너지 효율적인 생체모방 기반 노드 스케쥴링 알고리즘

손재현(Son Jae Hyun),변희정(Byun Hee Jung) 한국통신학회 2013 한국통신학회 학술대회논문집 Vol.2013 No.6

東京方言と長野縣方言アクセントの比較硏究

손재현 ( Jae Hyun Son ) 한국일어일문학회 2009 日語日文學硏究 Vol.68 No.1

옥외 환경 모니터링을 위한 센서노드 설계 및 구현

손재현(Jae-Hyun Son),조양행(Yang-Haeng Cho),김제홍(Je-Hong Kim),주영석(Young-Suk Joo),소선섭(Sun-Sup So) 대한전기학회 2007 대한전기학회 학술대회 논문집 Vol.2007 No.11

In this paper, we described a design and implementation of a sensor node for environmental monitoring. The main focus of design for sensor nodes is to isolate MCU for treating sensors from the RF module for considering various communication environment. The second is to make the interface between MCU and varity of sensor. In addition, we choose a narrow band communication module, cc1020, for the admittance of Korea government comminication law. We also use a uC/OS-Ⅱ as an operating system which is famous for 8bit MCUs. We showed that the communication performance is sufficient to use the communication module in a out-door environment through several experiments in that it is possible to transmit between 100m distance through experiments in a mountain.

서버리스 플랫폼에서 GPU 지원 및 인공지능 모델 추론에 적합한 함수 구조에 관한 연구

황동현 ( Dong-hyun Hwang ),김동민 ( Dongmin Kim ),최영윤 ( Young-yoon Choi ),한승호 ( Seung-ho Han ),전기만 ( Gi-man Jeon ),손재기 ( Jae-gi Son ) 한국정보처리학회 2019 한국정보처리학회 학술대회논문집 Vol.26 No.2

서버리스 프레임워크(Serverless Framework)는 마이크로서비스 아키텍처의 이론을 클라우드와 컨테이너를 기반으로 구현한 것으로 아마존의 AWS(Amazon Web Service)와 같은 퍼블릭 클라우드 플랫폼이 서비스됨에 따라 활용도 높아지고 있다. 하지만 현재까지의 플랫폼들은 GPU 와 같은 하드웨어의 의존성을 가진 인공지능 모델의 서비스에는 지원이 부족하다. 이에 본 논문에서는 컨테이너 기반의 오픈소스 서버리스 플랫폼을 대상으로 엔비디어-도커와 k8s-device-plugin 을 적용하여 GPU 활용이 가능한 서버리스 플랫폼을 구현하였다. 또한 인공지능 모델이 컨테이너에서 구동될 때 반복되는 가중치 로드를 줄이기 위한 구조를 제안한다. 본 논문에서 구현된 서버리스 플랫폼은 객체 검출모델인 SSD(Single Shot Multibox Detector) 모델을 이용하여 성능 비교 실험을 진행하였으며, 그 결과 인공지능 모델이 적용된 서버리스 플랫폼의 함수 응답 시간이 개선되었음을 확인하였다.