• 한국산업융합학회 논문집
• /
• 제23권6_2호
• /
• pp.1025-1032
• /
• 2020

In this paper, we propose a method to classify road weather conditions into rain, fog, and sun using a SVM (Support Vector Machine) classifier after extracting weather features from images acquired in real time using an optical sensor installed on a roadside post. A multi-dimensional weather feature vector consisting of factors such as image sharpeness, image entropy, Michelson contrast, MSCN (Mean Subtraction and Contrast Normalization), dark channel prior, image colorfulness, and local binary pattern as global features of weather-related images was extracted from road images, and then a road weather classifier was created by performing machine learning on 700 sun images, 2,000 rain images, and 1,000 fog images. Finally, the classification performance was tested for 140 sun images, 510 rain images, and 240 fog images. Overall classification performance is assessed to be applicable in real road services and can be enhanced further with optimization along with year-round data collection and training.

• 대한설비공학회지:설비저널
• /
• 제14권4호
• /
• pp.322-356
• /
• 1985

The outdoor design conditions for summer and winter are basic data required for determining the heating and cooling loads and HVAC equipment capacity. The latest study reported was based on the 1960's weather data, which is widely used by HVAC design engineers in Korea. The purpose of this paper is to update the outdoor design conditions for HVAC loads and equipments based on the weather data for the 1970's. The weather conditions of 24 sites, namely Sokcho, Chuncheon, Gangreung, Seoul, Inchon, Ulreungdo, Suweon, Seosan, Cheongju, Daejeon, Chupungryeong, Pohang, Gunsan, Daegu, Jeonju, Ulsan, Kwangju, Busan, Chungmu, Mokpo, Yeosu, Jeju, Seogwipo, and Jinju have been analyzed to calculate the outdoor design conditions. This analys is performed on the basis of TAC $1\%,\;TAC\;2.5\%,\;and\;TAC\;5\%$.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2024년도 춘계학술발표대회
• /
• pp.794-797
• /
• 2024

Despite advancements in deep learning, existing semantic segmentation models exhibit suboptimal performance under adverse weather conditions, such as fog or rain, whereas they perform well in clear weather conditions. To address this issue, much of the research has focused on making image or feature-level representations weather-independent. However, disentangling the style and content of images remains a challenge. In this work, we propose a novel fine-tuning method, 'freeze-n-update.' We identify a subset of model parameters that are weather-independent and demonstrate that by freezing these parameters and fine-tuning others, segmentation performance can be significantly improved. Experiments on a test dataset confirm both the effectiveness and practicality of our approach.

• 설비공학논문집
• /
• 제28권12호
• /
• pp.467-476
• /
• 2016

As the energy saving issues become one of the important global agenda, the building simulation method is generally used to predict the inside energy usage to establish the power-saving strategies. To foretell an accurate energy usage of a building, proper and typical weather data are needed. For this reason, typical weather data are fundamental in building energy simulations and among the meteorological factors, the solar irradiation is the most important element. Therefore, preparing solar irradiation is a basic factor. However, there are few places where the horizontal solar radiation in domestic weather stations can be measured, so the prediction of the solar radiation is needed to arrive at typical weather data. In this paper, four solar radiation prediction models were analyzed in terms of their applicability for domestic weather conditions. A total of 12 regions were analyzed to compare the differences of solar irradiation between measurements and the prediction results. The applicability of the solar irradiation prediction model for a certain region was determined by the comparisons. The results were that the Zhang and Huang model showed the highest accuracy (Rad 0.87~0.80) in most of the analyzed regions. The Kasten model which utilizes a simple regression equation exhibited the second-highest accuracy. The Angstrom-Prescott model is easily used, also by employing a plain regression equation Lastly, the Winslow model which is known for predicting global horizontal solar irradiation at any climate regions uses a daily integration equation and showed a low accuracy regarding the domestic climate conditions in Korea.

• 로봇학회논문지
• /
• 제17권3호
• /
• pp.245-254
• /
• 2022

Sensor dataset for autonomous driving is one of the essential components as the deep learning approaches are widely used. However, most driving datasets are focused on typical environments such as sunny or cloudy. In addition, most datasets deal with color images and lidar. In this paper, we propose a driving dataset with multi-spectral images and lidar in adverse weather conditions such as snowy, rainy, smoky, and dusty. The proposed data acquisition system has 4 types of cameras (color, near-infrared, shortwave, thermal), 1 lidar, 2 radars, and a navigation sensor. Our dataset is the first dataset that handles multi-spectral cameras in adverse weather conditions. The Proposed dataset is annotated as 2D semantic labels, 3D semantic labels, and 2D/3D bounding boxes. Many tasks are available on our dataset, for example, object detection and driveable region detection. We also present some experimental results on the adverse weather dataset.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제16권1호
• /
• pp.1-12
• /
• 2016

In this paper, we propose an approach that efficiently builds regional hazardous weather prediction models based on past weather data. Doing so requires finding the proper weather attributes that strongly affect hazardous weather for each region, and that requires a large number of experiments to build and test models with different attribute combinations for each kind of hazardous weather in each region. Using our proposed method, we reduce the number of experiments needed to find the correct weather attributes. Compared to the traditional method, our method decreases the number of experiments by about 45%, and the average prediction accuracy for all hazardous weather conditions and regions is 79.61%, which can help forecasters predict hazardous weather. The Korea Meteorological Administration currently uses the prediction models given in this paper.

• 한국태양에너지학회 논문집
• /
• 제37권1호
• /
• pp.25-38
• /
• 2017

Climate change and environmental pollution are becoming serious due to the use of fossil energy. For this reason, renewable energy systems are increasing, especially photovoltaic systems being more popular. The photovoltaic system has characteristics that are affected by ambient weather conditions such as insolation, outside temperature, wind speed. Particularly, it has been confirmed that the performance of the photovoltaic system decreases as the module temperature increases. In order to grasp the influence of the module temperature in advance, several researchers have proposed the prediction models on the module temperature. In this paper, we predicted the module temperature using the aforementioned prediction model on the basis of the weather conditions in Incheon, South Korea during July and August. The influence of weather conditions (i.e. insolation, outside temperature, and wind speed) on the accuracy of the prediction models was also evaluated using the standard statistical metrics such as RMSE, MAD, and MAPE. The results show that the prediction accuracy is reduced by 3.9 times and 1.9 times as the insolation and outside temperature increased respectively. On the other hand, the accuracy increased by 6.3 times as the wind speed increased.

• 한국태양에너지학회 논문집
• /
• 제39권6호
• /
• pp.41-54
• /
• 2019

The purpose of this study is to analyze the performance of solar thermal system according to regional weather conditions and feedwater temperature. The performance analysis of the system was carried out for the annual and winter periods in terms of solar fraction, collector efficiency and it's optimal degree. The system is simulated using TRNSYS program for 6 cities, Seoul, Incheon, Gangneung, Mokpo, Gwangju, and Ulsan. Simulation results prove that the solar fraction of the system varies greatly from region to region, depending on weather conditions and feedwater temperatures. Monthly average solar fraction for winter season from November to February, a time when heat energy is most required, indicated that the highest is 73.6% in Gangnueng and the lowest is 56.9% in Seoul. This is about 30% relative difference between the two cities. On the other hand, the collector efficiency of the system for all six cities was analyzed in the range between 40% and 42%, indicating small difference compare to the solar fraction. The annual average solar fraction is rated the highest at 40 collector degree, while monthly average solar fraction during winter season is rated at 60 degree.

• 한국화재소방학회:학술대회논문집
• /
• 한국화재소방학회 2008년도 추계학술논문발표회 논문집
• /
• pp.490-494
• /
• 2008

We simulated a forest fire which was occurred in Yangyang area on 2005 and compared a results between two different weather conditions(real weather condition and mean weather condition since 1968) using FARSITE, which is a forest fire spread simulator for preventing and predicting fire in USDA. And, we researched a problem in the transition for introducing, so we serve the basic method for prevention and attacking fire. In the result, severe weather condition on 2005 effected a forest fire behavior. The rate of spread under real weather condition was about 4 times faster than mean weather condition. Damaged area was about 10 time than mean weather condition. Therefore, Climate change will make a more sever fire season. As we will encounter to need for accurate prediction in near future, it will be necessary to predict a forest fire linked with future wether and fuel condition.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2001년도 춘계학술대회 논문집 전력기술부문
• /
• pp.190-193
• /
• 2001

In order to verify the efficiency or availability and stability of photovoltaic(PV) generation systems, huge system apparatuses are needed, in general, in which an actual size of solar panel, a type of converter system and some amount of load facilities should be installed in a particular location. It is also hardly possible to compare a Maximum Power Point Tracking (MPPT) control scheme with others under the same weather and load conditions in an actual PV generation system. The only and a possible way to bring above-mentioned problem to be solved is to realize a transient simulation scheme for PV generation systems using real weather conditions such as insolation and surface temperature of solar cell. The authors, in this paper, introduces a novel simulation method, which is based on a real-time digital simulator (RTDS), for PV generation systems under the real weather conditions. Firstly, VI characteristic equation of a solar cell is developed as an empirical formula and reconstructed in the RTDS system, then the real data of weather conditions are interfaced to the analogue inputs of the RTDS. The outcomes of the simulation demonstrate the effectiveness of the proposed simulation scheme in this paper. The results shows that the cost effective verifying for the efficiency or availability and stability of PV generation systems and the comparison research of various control schemes like MPPT under the same real weather conditions are possible.