• Nuclear Engineering and Technology
• /
• v.56 no.2
• /
• pp.558-567
• /
• 2024

Diagnostic support systems are being researched to assist operators in identifying and responding to abnormal events in a nuclear power plant. Most studies to date have considered single abnormal events only, for which it is relatively straightforward to obtain data to train the deep learning model of the diagnostic support system. However, cases in which multiple abnormal events occur must also be considered, for which obtaining training data becomes difficult due to the large number of combinations of possible abnormal events. This study proposes an approach to maintain diagnostic performance for multiple abnormal events by training a deep learning model with data on single abnormal events only. The proposed approach is applied to an existing algorithm that can perform feature selection and multi-label classification. We choose an extremely randomized trees classifier to select dedicated monitoring parameters for target abnormal events. In diagnosing each event occurrence independently, two-channel convolutional neural networks are employed as sub-models. The algorithm was tested in a case study with various scenarios, including single and multiple abnormal events. Results demonstrated that the proposed approach maintained diagnostic performance for 15 single abnormal events and significantly improved performance for 105 multiple abnormal events compared to the base model.

• Journal of Animal Science and Technology
• /
• v.63 no.1
• /
• pp.77-90
• /
• 2021

This study aimed to assess the impact of abnormal climate events on the production of Italian ryegrass (IRG), such as autumn low-temperature, severe winter cold and spring droughts in the central inland, southern inland and southern coastal regions. Seasonal climatic variables, including temperature, precipitation, wind speed, relative humidity, and sunshine duration, were used to set the abnormal climate events using principal component analysis, and the abnormal climate events were distinguished from normal using Euclidean-distance cluster analysis. Furthermore, to estimate the impact caused by abnormal climate events, the dry matter yield (DMY) of IRG between abnormal and normal climate events was compared using a t-test with 5% significance level. As a result, the impact to the DMY of IRG by abnormal climate events in the central inland of Korea was significantly large in order of severe winter cold, spring drought, and autumn low-temperature. In the southern inland regions, severe winter cold was also the most serious abnormal event. These results indicate that the severe cold is critical to IRG in inland regions. Meanwhile, in the southern coastal regions, where severe cold weather is rare, the spring drought was the most serious abnormal climate event. In particular, since 2005, the frequency of spring droughts has tended to increase. In consideration of the trend and frequency of spring drought events, it is likely that drought becomes a NEW NORMAL during spring in Korea. This study was carried out to assess the impact of seasonal abnormal climate events on the DMY of IRG, and it can be helpful to make a guideline for its vulnerability.

• Journal of the Korean Geotechnical Society
• /
• v.26 no.10
• /
• pp.69-79
• /
• 2010

To detect abnormal events in slopes, Principal Component Analysis (PCA) is applied to the slope that was collapsed during monitoring. Principal component analysis is a kind of statical methods and is called non-parametric modeling. In this analysis, principal component score indicates an abnormal behavior of slope. In an abnormal event, principal component score is relatively higher or lower compared to a normal situation so that there is a big score change in the case of abnormal. The results confirm that the abnormal events and collapses of slope were detected by using principal component analysis. It could be possible to predict quantitatively the slope behavior and abnormal events using principal component analysis.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.41 no.4
• /
• pp.294-300
• /
• 2008

Abnormal cooling of the Korean East Sea Water(KESW) in the East Sea before and after the 1982-1983 and 1997-1998 ENSO events is examined using bimonthly routine observation data from the National Fisheries Research and Development Institute of Korea for the period 1965 to 2002. The KESW, which occupies roughly a region between the Korean Peninsula and west of approximately $131^{\circ}E$, showed extreme cold-state years(1981 and 1996) prior to the two strongest ENSO events of the last half-century. Inter-annual bimonthly mean anomalies at 100 m in the KESW region were $-3.10^{\circ}C\;and\;-3.41^{\circ}C(SD=1.4^{\circ}C)$ in 1981 and 1996, respectively. These results suggest that extreme cooling of the KESW may be a prelude to very strong ENSO events through large-scale teleconnections.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.2
• /
• pp.197-206
• /
• 2011

The ability to detect online abnormal events in signals is essential in many real-world signal processing applications. In order to detect abnormal events, previously known algorithms require an explicit signal statistical model, and interpret abnormal events as statistical model abrupt changes. In general, maximum likelihood and Bayesian estimation theory to estimate well as detection methods have been used. However, the above-mentioned methods for robust and tractable model, it is not easy to estimate. More freedom to estimate how the model is needed. In this paper, we investigate a machine learning, descriptor-based approach that does not require a explicit descriptors statistical model, based on support vector machines are known to be robust statistical models and a sequential optimal algorithm online support vector machine is introduced.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.669-670
• /
• 2008

In this paper, we present a real-time system to detect abnormal events on gas pipes, based on the signals which are observed through the audio sensors attached on them. First, features are extracted from these signals so that they are robust to noise and invariant to the distance between a sensor and a spot at which an abnormal event like an attack on the gas pipes occurs. Then, a classifier is constructed to detect abnormal events using neural networks. It is a combination of two neural network models, a Gaussian mixture model and a multi-layer perceptron, for the reduction of miss and false alarms. The former works for miss alarm prevention and the latter for false alarm prevention. The experimental result with real data from the actual gas system shows that the proposed system is effective in detecting the dangerous events in real-time with an accuracy of 92.9%.

• Journal of the Korean Society of Industry Convergence
• /
• v.26 no.2_1
• /
• pp.217-223
• /
• 2023

With the rapid advancement of technologies, need for different research fields where this technology can be used is also increasing. One of the most researched topic in computer vision is object detection, which has widely been implemented in various fields which include healthcare, video surveillance and education. The main goal of object detection is to identify and categorize all the objects in a target environment. Specifically, methods of object detection consist of a variety of significant techniq ues, such as image processing and patterns recognition. Anomaly detection is a part of object detection, anomalies can be found various scenarios for example crowded places such as subway stations. An abnormal event can be assumed as a variation from the conventional scene. Since the abnormal event does not occur frequently, the distribution of normal and abnormal events is thoroughly imbalanced. In terms of public safety, abnormal events should be avoided and therefore immediate action need to be taken. When abnormal events occur in certain places, real time detection is required to prevent and protect the safety of the people. To solve the above problems, we propose a modified YOLOv5 object detection algorithm by implementing dilated convolutional layers which achieved 97% mAP50 compared to other five different models of YOLOv5. In addition to this, we also created a simple mobile application to avail the abnormal event detection on mobile phones.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.36 no.3
• /
• pp.34-42
• /
• 2013

In this study, we develop a method for analyzing business process based on the event-driven process chain (EPC) model. The method consists of five stages such as identifying abnormal events, finding causes for the abnormal events and problems caused by the abnormal events, making cause-and-effect chains, drawing root-cause map, and defining improvement areas. We illustrate how to apply the method with some examples for the domestic registered mail delivery process.

• Journal of KIISE
• /
• v.41 no.12
• /
• pp.1066-1074
• /
• 2014

In this paper, we present a visual analytics system that uses serial-correlation to detect an abnormal event in spatio-temporal data. Our approach extracts the topic-model from spatio-temporal tweets and then filters the abnormal event candidates using a seasonal-trend decomposition procedure based on Loess smoothing (STL). We re-extract the topic from the candidates, and then, we apply STL to the second candidate. Finally, we analyze the serial-correlation between the first candidates and the second candidate in order to detect abnormal events. We have used a visual analytic approach to detect the abnormal events, and therefore, the users can intuitively analyze abnormal event trends and cyclical patterns. For the case study, we have verified our visual analytics system by analyzing information related to two different events: the 'Gyeongju Mauna Resort collapse' and the 'Jindo-ferry sinking'.

• Environmental Engineering Research
• /
• v.16 no.3
• /
• pp.175-183
• /
• 2011

The climate of the $21^{st}$ century is likely to be significantly different from that of the 20th century because of human-induced climate change. An extreme weather event is defined as a climate phenomenon that has not been observed for the past 30 years and that may have occurred by climate change and climate variability. The abnormal climate change can induce natural disasters such as floods, droughts, typhoons, heavy snow, etc. How will the frequency and intensity of extreme weather events be affected by the global warming change in the $21^{st}$ century? This could be a quite interesting matter of concern to the hydrologists who will forecast the extreme weather events for preventing future natural disasters. In this study, we establish the extreme indices and analyze the trend of extreme weather events using extreme indices estimated from the observed data of 66 stations controlled by the Korea Meteorological Administration (KMA) in Korea. These analyses showed that spatially coherent and statistically significant changes in the extreme events of temperature and rainfall have occurred. Under the global climate change, Korea, unlike in the past, is now being affected by extreme weather events such as heavy rain and abnormal temperatures in addition to changes in climate phenomena.