• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.324-324
• /
• 2022

정확한 오염물질 예측은 기상학, 자연재해, 기후변화 연구 등 현장에서 필수적인 과제 중 하나이다. 주변 관측소에서 얻은 데이터를 사용하는 경우 모델 학습을 위한 불필요한 데이터로 인해 예측 결과에 왜곡 문제가 있을 수 있습니다. 따라서, 우리는 종합적인 대기질 지수 행동에 영향을 미치는 요인을 제공하는 최적의 데이터 소스를 찾기 위해 네트워크 방식을 사용했습니다. 본 연구에서는 2015년부터 2020년까지 우리나라의 6개 오염물질과 종합적인 대기질 지수 예측에 대한 네트워크 기법을 적용한 LSTM 및 DNN 모델을 적용하였다. 본 연구는 미세먼지(PM10), 초미세먼지(PM2.5), 오존(O3), 이산화황(SO2), 이산화질소(NO2), 일산화탄소(CO) 등 6가지 오염물질을 기반으로 종합적인 대기질 지수를 예측하는 2단계로 구성되어 있다. LSTM을 이용하여, 개별적으로 예측된 6가지 오염물질을 이용하여 DNN 모형을 이용하여 종합적인 대기질 지수를 예측한다. 6가지 오염물질에 대한 각 모델의 예측능력과 종합적인 대기질 지수 예측은 관측된 대기질 데이터와 비교하여 평가하였다. 본 연구는 심층신경망 모델과 네트워크 방식을 결합한 것이 높은 예측력을 제공함을 보여주었으며, 종합적인 대기질 지수 예측을 위한 최적의 모델로 선정되었다. 재난관리의 필요성이 증가함에 따라 네트워크 방식의 딥러닝 모델은 자연재해 피해를 줄이고 재난관리를 개선할 수 있는 충분한 잠재력을 가질 것으로 기대된다.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.12 no.4
• /
• pp.645-650
• /
• 2017

It is a system that controls temperature, humidity and fine-dust to make interior environment more comfortable for modern people who spend 90% of the time in indoor. In an experiment of finding which one of temperature and humidity influence more to discomfort index, for a fixed temperature of 21, discomfort index increased by 0.1 with a 1 change of humidity, and for a fixed humidity of 40, discomfort index increased by 1.2 with a 1 change of temperature. As a result, it was found that the temperature is more influential than the humidity to discomfort index. In an experiment of measuring communicational limitation of Bluetooth, the communication was possible for at most 30 meters without obstacles. With high obstacles like walls or steel bars, it was able to penetrate at most 2 obstacles and maximum distance which it can communicate was 10 meters for just one high molecule obstacle.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.8
• /
• pp.165-170
• /
• 2024

This study aims to verify the accuracy of the air quality management system in Yangju City using an artificial intelligence (AI) evaluation model. The consistency and reliability of fine dust data were assessed by comparing public data from the Ministry of Environment with data from Yangju City's air quality management system. To this end, we analyzed the completeness, uniqueness, validity, consistency, accuracy, and integrity of the data. Exploratory statistical analysis was employed to compare data consistency. The results of the AI-based data quality index evaluation revealed no statistically significant differences between the two datasets. Among AI-based algorithms, the random forest model demonstrated the highest predictive accuracy, with its performance evaluated through ROC curves and AUC. Notably, the random forest model was identified as a valuable tool for optimizing the air quality management system. This study confirms that the reliability and suitability of fine dust data can be effectively assessed using AI-based model performance evaluation, contributing to the advancement of air quality management strategies.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.490-490
• /
• 2022

국내 대하천 및 중·소규모 하천의 홍수터 공간을 활용하여 체육시설 및 공원 등과 같은 친수 시설물을 조성하여 친수공간으로 활용하고 있으며, 시민들의 친수활동 빈도는 증가추세에 있다. 특히 하천 내에서 수상 레크레이션 활동 등과 같은 다양한 친수활동이 증가하고 있으며, 하천친수에 관한 정보 수요가 급증하고 있으나 체계적인 공급은 미흡한 수준이다. 따라서 본 연구에서는 친수공간 조성 및 유지관리에 대한 측면과 친수공간에서의 쾌적한 친수활동을 위한 정보제공 목적으로 하천 친수공간에서의 복합모니터링을 이용한 친수활동 종합지수를 산정 방법을 개발하고자 하였다. 센서 기반의 시계열 데이터 구축을 위해 하천 수질, 수리인자의 복합모니터링을 진행하였다. 수리인자(수위, 유속, 수면폭 등)와 수질인자(탁도, Chl-a, pH 등), 기상학적 인자(자외선 지수, 미세먼지 등) 등급에 따른 허용기준을 설정하여 각 등급 별로 수리인자의 값을 0~1 사이 값인 소속도로 변환하여 소속도의 합성 및 친수활동 등급을 결정하였다. 최종적으로 수리, 수질, 기상 인자별 소속도 함수 산정을 통한 퍼지합성 이론 기반의 친수활동 종합지수를 산정하였다. 그리고 친수활동 종합지수를 예보하기 위한 모델 적용을 위한 방향성을 정립하였다.

• Proceedings of the Korea Air Pollution Research Association Conference
• /
• 2002.11a
• /
• pp.43-44
• /
• 2002

중국 서북건조 사막지역에서 봄철에 기온이 올라가면서 강한 한랭전선을 동반한 저기압이 발달할 때 그 전선 후면에 미세한 모래먼지가 바람에 휘감겨 올라가 황사가 발생한다. 발원지에서 생성된 황사는 약 30%가 발원지에서 재 침적되고, 20%는 주변지역으로 수송되며, 50%는 편서풍을 타고 1,500-2,000km이상의 거리를 이동하여 우리나라, 일본 및 하와이에까지 침적되는 것으로 알려져 있다. (중략)

• Journal of Environmental Science International
• /
• v.29 no.6
• /
• pp.623-631
• /
• 2020

This study analyzes the effect of negative air ions on the concentration of airborne particulate matter and evaluates the expected purification efficiency of open spaces for particulate matter by investigating the amount of negative air ions generated by plants. This study establishes a negative air ion generation treatment environment, plant environment, and control environment to measure the purification efficiency of particulate matter under the conditions of each, analyzing the expected purification efficiency by designing a particulate matter purification model. Results show that the amount of generated negative air ion according to environment was negative air ion generation treatment environment > plant environment > control environment; this order also applies to the particulate matter purification efficiency. Moreover, it took 65 min for the negative ion generation treatment environment, 90 min for the plant environment, and 240 min for the control environment to reach the standard expected purification efficiency of particulate matter concentration of 960 mg/㎥ for PM₁₀. For PM_2.5, with the designated maximum concentration of 700 mg/㎥, it took 60 min for the negative ion generation treatment environment, 80 min for the plant environment, and more than 240 min for the control environment. Based on these results, the expected purification efficiency compared to the control environment was quadrupled in the negative ion generation treatment environment and tripled in the plant environment on average.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.01a
• /
• pp.11-12
• /
• 2022

이 연구는 대기오염과 미세먼지의 각 성분이 질환에 미치는 영향에 대한 데이터만 존재한다면 어떠한 질환이든 위험도 예측 결과를 알 수 있는 것에 의미가 있다. 또한 기존의 대기정보에 따른 정보를 예상하는데 필요한 데이터 종류와 수가 많았으며 계산의 복잡성이 높았고 정보의 제공 범위가 넓었다. 하지만 이 연구는 과거 대기 데이터와 딥러닝을 통해서 낮은 비용으로 더욱 자세하게 유해질환 위험도를 예측하는 시스템을 구축하였다. 이 연구에서 구축한 시스템은 예측 결과 88.9%의 정확도를 보였다. 이 시스템은 입력되는 데이터의 정보에 따라 세분화된 지역의 대기환경 정보 또한 파악 가능하며 그 과정이 매우 간편하고 유용하다. 이 시스템은 공기질 예측을 위해 유용하게 사용될 수 있을 것이라고 사료된다.

• Journal of the Korean earth science society
• /
• v.44 no.3
• /
• pp.196-209
• /
• 2023

The Nishinoshima volcano, located 940 km south of Tokyo, experienced an eruption from June to August 2020. The volcanic gas and ash from the eruption of Nishinoshima that occurred at the end of July 2020 was reported to have the potential to affect the Korean Peninsula. In this study, we used Ash3D, a numerical simulation program for volcanic ash dispersion, to investigate the eruption that occurred at 0:00 local time on July 28, 2020, with a volcanic explosivity index of three. The results showed that the volcanic ash cloud reached Okinawa on the morning of July 30, carried by an east wind. It then moved northward and reached Jeju Island on August 1, eventually circulating in a clockwise direction and reaching southern part of the Korean Peninsula on August 2. The concentration of Particulate Matter 10 (PM₁₀), measured at the Jeju Gosan Meteorological Observatory in Jeju Island, increase from August 1. A similar increase in PM₁₀ concentration was observed at the Gudeok Mountain Weather Station in Busan from August 2. These findings suggested that eruption of the Nishinoshima volcano had an impact on the fine dust concentrations at Jeju Island and southern part of the Korean Peninsula.

• Environmental and Resource Economics Review
• /
• v.25 no.4
• /
• pp.545-564
• /
• 2016

Recently as fine and ultra fine particles become major environmental issues in Korea, it is very important to develop effective solutions to air pollution. Accordingly this study aims at detecting causes of air pollution by using models and examining if diesel price increases contribute to reduction of diesel consumption and air pollution. TSP, PM10, $NO_X$, $SO_X$, CO, and VOC are included as major air pollutants. As a result, we found invert U shape curve between pollution and income for all air pollutants except CO. Consumer price index, coal power capacity, diesel consumption, frequency of yellow dust, number of natural gas buses, number of transport business, annual average temperature, number of manufacturing businesses are also influential in explaining causes of air pollution. As diesel price increases by 1%, air pollutants decline between 0.07~0.12% in the short run. Simultaneously, the additional revenue from increases in diesel prices might be transferred to support expansion of biofuel market. Also, stronger policy should be developed to mitigate the current air pollution problem.

• Tuberculosis and Respiratory Diseases
• /
• v.55 no.6
• /
• pp.570-578
• /
• 2003

Background : Ambient particles during Asian dust events are usually sized less than $10{\mu}m$, known to be associated with the adverse effects on the general populations. But, there has been no considerable evidence linking these particles to the adverse effects on airways. The objectives of this study was to investigate the possible adverse effects of Asian dust events on respiratory function and symptoms in subjects with bronchial asthma. Patients and Methods : From march to June 2002, Asthmatic patients who were diagnosed with bronchial challenge test or bronchodilator response were enrolled. We divided them into three groups; mild, moderate, and severe, according to the severity. Subjects with other organ insufficiency such as heart, kidney, liver, and malignancy were excluded. All patients completed twice daily diaries and recorded peak flow rate, respiratory symptom, and daily activity. Daily and hourly mean pollutant levels of particulate matter < $10{\mu}m$ in diameter($PM_{10}$), nitrogen dioxide($NO_2$), sulphur dioxide($SO_2$), ozone($O_3$) and carbon monoxide(CO) were measured at the 10 different monitoring sites. Results : Dust events occured 14 times during the study period. Daily averages of 4 air pollutant were measured with an increased level of $PM_{10}$, decreased level of $NO_2$ and $SO_2$, and no change in CO during dust days compared to those during control days. An increase in $PM_{10}$ concentration was associated with an increase of subjects with PEF variability of >20% (p<0.05), night time symptom(p<0.05), and a decrease in mean PEF (p<0.05), which were calculated by the longitudinal data analysis. Otherwise, there was no association between $PM_{10}$ level and bronchodialtor inhaler, and daytime respiratory symptoms. Conclusion : This study shows evidence that ambient air pollution, especially $PM_{10}$, during Asian dust events, could be one of the many aggravating factors at least in patients with airway diseases. This data can be used as a primary source to set up a new policy on air environmental control and to evaluate the safety of air pollution index. We also expect that this research will help identify precise components of dust, which are more linked to the adverse effects.