• Nuclear Engineering and Technology
• /
• 제55권5호
• /
• pp.1630-1643
• /
• 2023

The correct situation awareness (SA) of operators is important for managing nuclear power plants (NPPs), particularly in accident-related situations. Among the three levels of SA suggested by Ensley, Level 3 SA (i.e., projection of the future status of the situation) is challenging because of the complexity of NPPs as well as the uncertainty of accidents. Hence, several prediction methods using artificial intelligence techniques have been proposed to assist operators in accident prediction. However, these methods only predict short-term plant status (e.g., the status after a few minutes) and do not provide information regarding the uncertainty associated with the prediction. This paper proposes an algorithm that can predict the multivariate and long-term behavior of plant parameters for 2 h with 120 steps and provide the uncertainty of the prediction. The algorithm applies bidirectional long short-term memory and an attention mechanism, which enable the algorithm to predict the precise long-term trends of the parameters with high prediction accuracy. A conditional variational autoencoder was used to provide uncertainty information about the network prediction. The algorithm was trained, optimized, and validated using a compact nuclear simulator for a Westinghouse 900 MWe NPP.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2002년도 추계종합학술대회
• /
• pp.800-803
• /
• 2002

The MONOS capacitor are fabricated to investigate the carrier trapping due to Fowler-Nordheim tunneling injection. The carrier trapping in scaled multi-dielectric(ONO) depends on the nitride and Op oxide thickness under Fowler_Nordheim tunneling injection. Carriers captured at nitride film could not escape from nitride to gate, but be captured at top oxide and nitride interface traps because of barrier height of top oxide. Therefore, it is expected that the MONOS memory devices using multi dielectric films enhance memory effect and have a long memory retention characteristic.

• 한국지하수토양환경학회:학술대회논문집
• /
• 한국지하수토양환경학회 2004년도 총회 및 춘계학술발표회
• /
• pp.451-454
• /
• 2004

The study aims to assess the quality of bank filtrate in relation to streamflow and physico-chemical properties of the stream. Turbidity, pH, temperature and dissolved oxygen (DO) of Nakdong River and riverbank filtrate were statistically analyzed. The physico-chemical properties of riverbank filtrate were measured from irregularly different seven pumping wells every day. Autocorrelation analyses were conducted to the qualities of stream water and bank filtrated water. Temperature, pH and DO of streamflow shows strong linearity and long memory effect, indicating the effect of seasonal air temperature and rainy season. Temperature of riverbank filtrate shows weak linearity and weak memory, indicating differently from the trend of stream temperature. Turbidity of steramflow shows strong linearity and long memory effect, while turbidity of riverbank filtrate indicates weak linearity and weak memory. Cross-correlation analysis shows low relation between turbidity, pH, temperature and DO of riverbank filtrate and those of streamflow. Turbidity of streamflow was largely affected by the streamflow rate, showing a similar trend with autocorrelation function of streamflow rate. The turbidity of riverbank filtrate has a lag time of 25 hours. This indicates that turbidity of streamflow in a dry season has very low effect on the turbidity of riverbank filtrate, and a high turbidity of the stream in a rainy season has a fairly low effect on the turbidity of riverbank filtrate.

• 대한의생명과학회지
• /
• 제28권4호
• /
• pp.211-222
• /
• 2022

After more than two years of efforts to end the corona pandemic, a gradual recovery is starting in countries with high vaccination rates. Easing public health policies for a full-fledged post-corona era, such as lifting the mandatory use of outdoor mask and quarantine measures in entry have been considered in Korea. However, the continuous emergence of new variants of SARS-CoV-2 and limitations in vaccine efficacy still remain challenging. Fortunately, T cells and memory T cells, which are key components of adaptive immunity appear to contribute substantially in COVID-19 control. SARS-CoV-2 specific CD4⁺/CD8⁺ T cells are induced by natural infection or vaccination, and rapid induction and activation of T cells is mainly associated with viral clearance and attenuated clinical severity. In addition, T cell responses induced by recognition of a wide range of epitopes were minimally affected and conserved against the highly infectious subsets of omicron variants. Polyfunctional SARS-CoV-2 specific T cell memory including stem cell-like memory T cells were also developed in COVID-19 convalescent patients, suggesting long lasting protective T cell immunity. Thus, a robust T-cell immune response appears to serve as a reliable and long-term component of host protection in the context of reduced efficacy of humoral immunity and persistent mutations and/or immune escape.

• 마이크로전자및패키징학회지
• /
• 제20권2호
• /
• pp.1-9
• /
• 2013

Recently flexible electronic devices have attracted a great deal of attention because of new application possibilities including flexible display, flexible memory, flexible solar cell and flexible sensor. In particular, development of flexible memory is essential to complete the flexible integrated systems such as flexible smart phone and wearable computer. Research of flexible memory has primarily focused on organic-based materials. However, organic flexible memory has still several disadvantages, including lower electrical performance and long-term reliability. Therefore, emerging research in flexible electronics seeks to develop flexible and stretchable technologies that offer the high performance of conventional wafer-based devices as well as superior flexibility. Development of flexible memory with inorganic silicon materials is based on the design principle that any material, in sufficiently thin form, is flexible and bendable since the bending strain is directly proportional to thickness. This article reviews progress in recent technologies for flexible memory and flexible electronics with inorganic silicon materials, including transfer printing technology, wavy or serpentine interconnection structure for reducing strain, and wafer thinning technology.

• Biomolecules & Therapeutics
• /
• 제23권3호
• /
• pp.251-260
• /
• 2015

Propofol is an anesthetic agent that gained wide use because of its fast induction of anesthesia and rapid recovery post-anesthesia. However, previous studies have reported immediate neurodegeneration and long-term impairment in spatial learning and memory from repeated neonatal propofol administration in animals. Yet, none of those studies has explored the sex-specific long-term physical changes and behavioral alterations such as social (sociability and social preference), emotional (anxiety), and other cognitive functions (spatial working, recognition, and avoidance memory) after neonatal propofol treatment. Seven-day-old Wistar-Kyoto (WKY) rats underwent repeated daily intraperitoneal injections of propofol or normal saline for 7 days. Starting fourth week of age and onwards, rats were subjected to behavior tests including open-field, elevated-plus-maze, Y-maze, 3-chamber social interaction, novel-object-recognition, passive-avoidance, and rotarod. Rats were sacrificed at 9 weeks and hippocampal protein expressions were analyzed by Western blot. Results revealed long-term body weight gain alterations in the growing rats and sex-specific impairments in spatial (female) and recognition (male) learning and memory paradigms. A markedly decreased expression of hippocampal NMDA receptor GluN1 subunit in female- and increased expression of AMPA GluR1 subunit protein expression in male rats were also found. Other aspects of behaviors such as locomotor activity and coordination, anxiety, sociability, social preference and avoidance learning and memory were not generally affected. These results suggest that neonatal repeated propofol administration disrupts normal growth and some aspects of neurodevelopment in rats in a sex-specific manner.

• Communications for Statistical Applications and Methods
• /
• 제26권5호
• /
• pp.497-506
• /
• 2019

Forecasting the U.S. employment level is made using machine learning methods of the artificial neural network: deep neural network, long short term memory (LSTM), gated recurrent unit (GRU). We consider the big data of the federal reserve economic data among which 105 important macroeconomic variables chosen by McCracken and Ng (Journal of Business and Economic Statistics, 34, 574-589, 2016) are considered as predictors. We investigate the influence of the two statistical issues of the dimension reduction and time series differencing on the machine learning forecast. An out-of-sample forecast comparison shows that (LSTM, GRU) with differencing performs better than the autoregressive model and the dimension reduction improves long-term forecasts and some short-term forecasts.

• The Plant Pathology Journal
• /
• 제38권4호
• /
• pp.395-402
• /
• 2022

To predict rice blast, many machine learning methods have been proposed. As the quality and quantity of input data are essential for machine learning techniques, this study develops three artificial neural network (ANN)-based rice blast prediction models by combining two ANN models, the feed-forward neural network (FFNN) and long short-term memory, with diverse input datasets, and compares their performance. The Blast_Weathe long short-term memory r_FFNN model had the highest recall score (66.3%) for rice blast prediction. This model requires two types of input data: blast occurrence data for the last 3 years and weather data (daily maximum temperature, relative humidity, and precipitation) between January and July of the prediction year. This study showed that the performance of an ANN-based disease prediction model was improved by applying suitable machine learning techniques together with the optimization of hyperparameter tuning involving input data. Moreover, we highlight the importance of the systematic collection of long-term disease data.

• 한국빅데이터학회지
• /
• 제5권1호
• /
• pp.1-9
• /
• 2020

이 연구는 2018년도 정부(농림축산식품부)의 재원으로 농림식품기술기획평가원 지원을 받아 수행된 연구이다. 최근 시계열 및 텍스트 마이닝에서 활발히 사용되는 모델은 딥러닝(Deep Learning) 모델 구조를 활용한 LSTM(Long Short-Term Memory models) 모델이다. LSTM 모델은 RNN의 BPTT(Backpropagation Through Time) 과정에서 발생하는 Long-Term Dependency Problem을 해결하기 위해 등장한 모델이다. LSTM 모델은 가변적인 Sequence data를 활용하여 예측하는 문제를 굉장히 잘 해결했고, 지금도 널리 사용되고 있다. 본 논문 연구에서는 KT가 제공하는 CDR(Call Detailed Record) 데이터를 활용하여 바이러스와 밀접한 관계가 있을 것으로 예측되는 사람의 이동 경로를 파악하였다. 해당 사람의 경로를 활용하여 LSTM 모델을 학습시켜 이동 경로를 예측한 결과를 소개한다. 본 연구 결과를 활용하여 HPAI가 전파되는 경로를 예측하여 방역에 중점을 둘 경로 또는 지역을 선정해 HPAI 확산을 줄이는 데 이용될 수 있을 것이다.

• 대한토목학회논문집
• /
• 제44권1호
• /
• pp.11-17
• /
• 2024

제진 구조는 댐퍼라는 장치를 구조물에 장착시켜 지진에너지를 소산하는 내진설계이다. 지진피해를 저감하고자 하는 연구가 성행하고 있는 가운데 제진 구조는 댐퍼의 재료, 형상을 변경함으로써 기술을 발전시켜왔다. 하지만 댐퍼의 특성상 에너지를 소산하기 위해 재료에 발생하는 소성변형은 피할 수 없는 한계가 있다. 따라서 본 연구에서는 발생한 변형를 스스로 회복할 수 있는 초탄성 형상기억합금(Superelastic shape memory alloy, SSMA)을 활용하여 반영구적으로 사용할 수 있고 추가적인 긴장력을 적용하여 구조적 성능을 향상한 장수명 댐퍼를 제안하였다. 장수명 댐퍼의 거동 특성 분석을 위해 재료, 와이어 직경, 긴장력 유무의 설계 변수에 따라 유한요소해석을 진행하였고 응답 거동을 도출하여 하중 저항, 에너지 소산, 잔류변위 등의 특성을 분석하여 장수명 댐퍼의 성능적 우수성을 입증하였다.