http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
입력 데이터 해상도에 따른 심층학습 알고리즘의 기상 변수 예측 정확도 평가
서상익,이창준,기서진 한국환경기술학회 2022 한국환경기술학회지 Vol.23 No.1
We evaluated the performance of deep learning algorithms predicting air temperature in different time steps. Three different data sets were compiled at various time intervals covering days, hours, and minutes for three separate months (i.e., January, July, and November 2021) in two monitoring stations (i.e., one in Seoul 108 and the other in Jinju 192) from the Korea Meteorological Administration. Those data sets divided into 70 % for training and 30 % for testing were provided as inputs to two popular algorithms, the multi layer perceptron (MLP) and long short-term memory (LSTM). Our results showed that the MLP algorithm exhibited superior prediction performance for data recorded at one-minute intervals rather than those updated hourly or daily. In addition, the MLP algorithm was found to work best for data with seasonality. The predictive accuracy was, however, slightly lower for the MLP algorithm than for the LSTM algorithm which yielded error rates as low as 0.04 in terms of the mean absolute error. All these results implied that the use of high-frequency data played an important role in improving the performance of deep learning as well as the proposed methodology could be used to prioritize candidate algorithms with input data (resolution) for prediction of weather variables.
서상익 ( Sang-ik Suh ),박정호 ( Jung-ho Park ),서현효 ( Hyun-hyo Suh ) 한국환경기술학회 2024 한국환경기술학회지 Vol.25 No.1
For effective treatment of recalcitrant industrial wastewater, microbial augmentation J30, which is composed of Pseudomonas sp. GT21, Bacillus sp. KN27, Acinetobacter sp. KN11, and Neisseria sp. KN13, decomposing strains that show high activities on various organic substances and aromatic compounds, were prepared. As a method to control the solid suspension of industrial wastewater, the culture solutions of A and B, which produce bioflocculants, were applied to wastewater treatment. A microbial augmentation for treating laundry wastewater containing TCE and PCE were prepared by mixing 4 types of strains used to treat paper mill wastewater, 6 types of strains prepared by formulating culture solutions of Flavobacterium sp. GN18 and Acalgenes sp. GN23, which are richloroethylene and pentachloroethylene decomposing strains. It was named microbial augmentation TP32. These culture solutions of strain groups were added to treat paper mill wastewater containing phenolic chlorine compounds and laundry wastewater containing chlorine compounds efficiently. The removal efficiencies of the treatment groups, for TOC in paper mill wastewater were 81.9 % and 73.2 %, respectively. The TOC removal efficiency in paper mill wastewater was 86.5 % in the mixed strains treatment group, where decomposing and flocculating strains were added simultaneously. Changes in T-N and T-P showed a removal efficiency of more than 70 % in all treatments using decomposing, flocculating, and mixed strains. Still, the mixed strain treatment showed the highest removal efficiency. The phenolic chlorine compound, 3-chlorophenol removal efficiency was 85.1 % in the mixed strain treatment group. The removal efficiency of TCE from laundry wastewater showed a removal rate of 85.3 % in 42 hours of incubation when only decomposing strains were added and a removal rate of more than 90 % in 32 hours of incubation when mixed strains were added. The PCE removal rate has been decreased over time of incubation in the decomposing strain treatment and mixed strain treatment, respectively. For effective treatment of recalcitrant industrial wastewater, microbial augmentation J30, which is composed of Pseudomonas sp. GT21, Bacillus sp. KN27, Acinetobacter sp. KN11, and Neisseria sp. KN13, decomposing strains that show high activities on various organic substances and aromatic compounds, were prepared. As a method to control the solid suspension of industrial wastewater, the culture solutions of A and B, which produce bioflocculants, were applied to wastewater treatment. A microbial augmentation for treating laundry wastewater containing TCE and PCE were prepared by mixing 4 types of strains used to treat paper mill wastewater, 6 types of strains prepared by formulating culture solutions of Flavobacterium sp. GN18 and Acalgenes sp. GN23, which are richloroethylene and pentachloroethylene decomposing strains. It was named microbial augmentation TP32. These culture solutions of strain groups were added to treat paper mill wastewater containing phenolic chlorine compounds and laundry wastewater containing chlorine compounds efficiently. The removal efficiencies of the treatment groups, for TOC in paper mill wastewater were 81.9 % and 73.2 %, respectively. The TOC removal efficiency in paper mill wastewater was 86.5 % in the mixed strains treatment group, where decomposing and flocculating strains were added simultaneously. Changes in T-N and T-P showed a removal efficiency of more than 70 % in all treatments using decomposing, flocculating, and mixed strains. Still, the mixed strain treatment showed the highest removal efficiency. The phenolic chlorine compound, 3-chlorophenol removal efficiency was 85.1 % in the mixed strain treatment group. The removal efficiency of TCE from laundry wastewater showed a removal rate of 85.3 % in 42 hours of incubation when only decomposing strains were added and a removal rate of more than 90 % in 32 hours of incubation when mixed strains were added. The PCE removal rate has been decreased over time of incubation in the decomposing strain treatment and mixed strain treatment, respectively.
농업분야의 노출 위해성 평가 모델 개발을 위한 입력 자료 및 시스템 요구사항 분석
서상익 ( Sang-ik Suh ),조경철 ( Gyeongcheol Jo ),기서진 ( Seo Jin Ki ) 한국환경농학회 2021 한국환경농학회 학술대회집 Vol.2021 No.-
광역 스케일의 노출 위해성 평가 모델은 세부적인 물리기반 모델에 비해 (적은 수의 입력 자료를 활용하여) 넓은 지역의 위해성을 보다 효율적을 평가할 수 있고, 특히 위해성이 상대적으로 높은 지역을 빠른 시간에 판별할 수 있는 장점이 있다. 본 연구에서는 농업활동 및 농업용수 유래 유해물질 노출에 따른 (오프라인 기반의) 위해성 평가 모델 개발하고 (최종적으로) 온라인 위해성 평가 시스템을 구축하기 위해 수행되었으며, 이를 위한 예비 연구로 위해성 평가 모델에 필요한 입력 자료 검토하고 오프라인 시스템 요구 사항을 우선적으로 분석하였다. 유럽과 미국 등의 선진국에서 위해성 평가 모델을 검토한 결과 위해성 평가 방식이 상이함을 알 수 있었으며, 이에 따라 요구되는 입력 자료 수준이 상이함을 알 수 있었다. 다만, 위해성 평가에 공통적으로 요구되는 대상 농약에 대한 화학적 특성은 동일한 것으로 판단되어 국내 노출 위해성 평가 모델 개발에 활용할 수 있을 것으로 판단된다. 노출 위해성 평가 모델에 적용하고자 하는 총 190개의 후보 대상 농약 중 일차적으로 53개 오염물질의 특성 DB가 구축되었으며, 향후 추가적인 문헌검토 작업을 통해 나머지 오염물질의 특성 DB가 구축될 경우 국내 노출 위해성 모델에 활용될 수 있을 것으로 판단된다. 오프라인 기반의 노출 위해성 평가 모델을 지리정보시스템(Geographic Information System, GIS)에 탑재할 방안을 검토한 결과, 다양한 오픈 소스 기반의 GIS 소프트웨어 중 최근 환경 및 생태 등의 분야에 활용도가 높은 QGIS 프로그램이 적합한 것으로 판단된다. 또한, 노출 위해성 평가 모델은 Python 콘솔. Python 프로세싱 도구, Python 플러그인 등의 방법을 사용하여 QGIS 소프트웨어 내 탑재가 가능할 것으로 평가되었다.