http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
학교 교실 실내공기질 시계열 데이터의 순환신경망 기반 예측모델 개발
태경응(Kyung Eung Tae),백창인(Chang In Baek),이경순(Kyung Soon Lee),이정재(Jurng Jae Yee) 대한설비공학회 2022 대한설비공학회 학술발표대회논문집 Vol.2022 No.11
Predicting indoor air quality in indoor spaces can alert residents to improve indoor air quality before the indoor air quality deteriorates, which can greatly help indoor residents comfort, health, safety and hygiene. We measured indoor air quality (CO2 concentration, PM2.5, temperature, humidity, TVOCs) in school classrooms and developed CO₂ and PM2.5 concentration prediction models using Recurrent Neural Networks (RNN) and validated their feasibility. To predict indoor air quality in the next 10 minutes, we investigated how much historical data is good to use. Studies have shown that models using LSTM recurrent neural networks have well predicted indoor air quality in classrooms, but have been burdened with time or memory spent on training models, and considering their practicality in the field, lighter prediction models are needed.
미세먼지 대응 고효율 필터가 적용된 혹한기 열 회수형 환기장치 개발
태경응(Kyung Eung Tae),명노훈(No Hoon Myeong),강경모(Kyung Mo Kang),이윤규(Yun Gyu Lee) 대한설비공학회 2017 대한설비공학회 학술발표대회논문집 Vol.2017 No.6
Energy recovery ventilator (ERV) with application of pre-heater, by-pass and high efficient filter provides an opportunity to eliminate condensation in ventilator. However, building energy consumption increases for operating the heater. We applied to inner return damper rising air temperature without pre-heater. In the experiment, the condensation was not detected in -20℃. The highest CO2 concentrations measured 991ppm operating return damper. When the high efficiency air filter was applied, power-consumption increased due to increasing the static pressure and the indoor PM10 concentrations decreased 28.4% less than without the filter. Applying to by-pass improve energy consumptions and indoor air quality.
반도체 클린룸용 외기공조시스템의 수분무 가습을 이용한 에너지절감에 관한 연구
송원일,김기철,유경훈,신대건,태경응,김용식,박덕준,Song, Won-Il,Kim, Ki-Cheol,Yoo, Kyung-Hoon,Shin, Dae-Kun,Tae, Kyung-Eung,Kim, Yong-Sik,Park, Dug-Jun 한국입자에어로졸학회 2017 Particle and Aerosol Research Vol.13 No.2
In recent large-scale semiconductor manufacturing cleanrooms, the energy consumption in outdoor air conditioning (OAC) systems to heat, humidify, cool and dehumidify outdoor air(OA) represents about 40~50 % of the total cleanroom power consumption required to maintain cleanroom environment. Therefore, the assessment of energy consumption in outdoor air conditioning systems is essential for reducing the outdoor air conditioning load for a cleanroom. In the present study, an experiment with an outdoor air flow rate of $1,000m^3/h$ was conducted to compare the energy consumption in steam humidification, simple air washer, exhaust air heat recovery type air washer and dry cooling coil(DCC) return water heat recovery type air washer OAC systems. Besides, a numerical analysis was carried out to evaluate the annual energy consumption of the aforementioned four OAC systems. It was shown that the simple air washer, exhaust air heat recovery type air washer and DCC return water heat recovery type air washer OAC systems using water spray humidification were more energy-efficient than the steam humidification OAC system. Furthermore the DCC return water heat recovery type air washer OAC system was the most energy-efficient.
백창인(Chang In Baek),태경응(Kyung Eung Tae),이경순(Kyung Soon Lee),이정재(Jurng Jae Yee) 대한설비공학회 2021 대한설비공학회 학술발표대회논문집 Vol.2021 No.6
Predicting indoor air quality in indoor spaces can alert residents to improve indoor air quality before the indoor air quality deteriorates, which can greatly help indoor residents comfort, health, safety and hygiene. We measured indoor air quality (CO2 concentration, PM2.5, temperature, humidity, TVOCs) in school classrooms and developed CO₂ and PM2.5 concentration prediction models using Recurrent Neural Networks (RNN) and validated their feasibility. To predict indoor air quality in the next 10 minutes, we investigated how much historical data is good to use. Studies have shown that models using LSTM recurrent neural networks have well predicted indoor air quality in classrooms, but have been burdened with time or memory spent on training models, and considering their practicality in the field, lighter prediction models are needed.