• Earthquakes and Structures
• /
• v.22 no.4
• /
• pp.401-419
• /
• 2022

A novel hybrid extreme machine learning-multiverse optimizer (ELM-MVO) framework is proposed to predict the liquefaction phenomenon in seismically excited tunnel lining inside the sand lens. The MVO is applied to optimize the input weights and biases of the ELM algorithm to improve its efficiency. The tunnel located inside the liquefied sand lens is also evaluated under various near- and far-field earthquakes. The results demonstrate the superiority of the proposed method to predict the liquefaction event against the conventional extreme machine learning (ELM) and artificial neural network (ANN) algorithms. The outcomes also indicate that the possibility of liquefaction in sand lenses under far-field seismic excitations is much less than the near-field excitations, even with a small magnitude. Hence, tunnels designed in geographical areas where seismic excitations are more likely to be generated in the near area should be specially prepared. The sand lens around the tunnel also has larger settlements due to liquefaction.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.11a
• /
• pp.1004-1007
• /
• 2020

본 논문에서는 청각 장애인이 가정에서 듣지 못해 발생하는 낭비와 위험을 미리 예방하기 위하여 가정에서 현재 발생하고 있는 소리를 알려주는 시스템을 구현하였다. 무지향성 마이크로 일상 소리 감지 후 음향 데이터에서 Mel-Spectogram 특징 벡터를 추출하여 Convolutional Neural Network(CNN) 모델의 Resnet 알고리즘을 진행한다. 서버에서 소리에 대한 분석을 진행한 후 그 결과를 안드로이드에서 실시간으로 5 초마다 확인하여 사용자에게 알림 서비스를 제공한다. 이를 통해 낭비를 줄이고 위험에 대처할 수 있게 한다. 청각 장애인의 소리에 대한 접근성을 다양한 측면으로 고려해야 한다는 사회적 인식을 확산시키고자 한다.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.38 no.5
• /
• pp.671-683
• /
• 2018

The flood damage caused by heavy rains in urban watershed is increasing, and, as evidenced by many previous studies, urban flooding usually exceeds the water capacity of drainage networks. The flood on the area which considerably urbanized and densely populated cause serious social and economic damage. To solve this problem, deterministic and probabilistic studies have been conducted for the prediction flooding in urban areas. However, it is insufficient to obtain lead times and to derive the prediction results for the flood volume in a short period of time. In this study, IDNN, TDNN and NARX were compared for real-time flood prediction based on urban runoff analysis to present the optimal real-time urban flood prediction technique. As a result of the flood prediction with rainfall event of 2010 and 2011 in Gangnam area, the Nash efficiency coefficient of the input delay artificial neural network, the time delay neural network and nonlinear autoregressive network with exogenous inputs are 0.86, 0.92, 0.99 and 0.53, 0.41, 0.98 respectively. Comparing with the result of the error analysis on the predicted result, it is revealed that the use of nonlinear autoregressive network with exogenous inputs must be appropriate for the establishment of urban flood response system in the future.

• Journal of Korean Society of Disaster and Security
• /
• v.14 no.4
• /
• pp.17-27
• /
• 2021

Streamflow prediction is a very vital disaster mitigation approach for effective flood management and water resources planning. Lately, torrential rainfall caused by climate change has been reported to have increased globally, thereby causing enormous infrastructural loss, properties and lives. This study evaluates the contribution of rainfall to streamflow prediction in normal and peak rainfall scenarios, typical of the recent flood at Piney Resort in Vernon, Hickman County, Tennessee, United States. Daily streamflow, water level, and rainfall data for 20 years (2000-2019) from two USGS gage stations (03602500 upstream and 03599500 downstream) of the Piney River watershed were obtained, preprocesssed and fitted with Long short term memory (LSTM) model. Tensorflow and Keras machine learning frameworks were used with Python to predict streamflow values with a sequence size of 14 days, to determine whether the model could have predicted the flooding event in August 21, 2021. Model skill analysis showed that LSTM model with full data (water level, streamflow and rainfall) performed better than the Naive Model except some rainfall models, indicating that only rainfall is insufficient for streamflow prediction. The final LSTM model recorded optimal NSE and RMSE values of 0.68 and 13.84 m³/s and predicted peak flow with the lowest prediction error of 11.6%, indicating that the final model could have predicted the flood on August 24, 2021 given a peak rainfall scenario. Adequate knowledge of rainfall patterns will guide hydrologists and disaster prevention managers in designing efficient early warning systems and policies aimed at mitigating flood risks.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.37 no.3
• /
• pp.163-171
• /
• 2024

Structural health monitoring for ships and offshore structures is important in various aspects. Ships and offshore structures are continuously exposed to various environmental conditions, such as waves, wind, and currents. In the event of an accident, immense economic losses, environmental pollution, and safety problems can occur, so it is necessary to detect structural damage or defects early. In this study, structural response data of multi-linked floating offshore structures under various wave load conditions was calculated by performing fluid-structure coupled analysis. Furthermore, the order reduction method with distortion base mode was applied to the structures for predicting the structural response by using the results of numerical analysis. The distortion base mode order reduction method can predict the structural response of a desired area with high accuracy, but prediction performance is affected by sensor arrangement. Optimization based on a genetic algorithm was performed to search for optimal sensor arrangement and improve the prediction performance of the distortion base mode-based reduced-order model. Consequently, a sensor arrangement that predicted the structural response with an error of about 84.0% less than the initial sensor arrangement was derived based on the root mean squared error, which is a prediction performance evaluation index. The computational cost was reduced by about 8 times compared to evaluating the prediction performance of reduced-order models for a total of 43,758 sensor arrangement combinations. and the expected performance was overturned to approximately 84.0% based on sensor placement, including the largest square root error.

• Atmosphere
• /
• v.19 no.1
• /
• pp.37-51
• /
• 2009

Kinematic characteristics of a heavy rainfall event occurred in Changma front are analyzed using synoptic weather charts, satellite imagery and NCEP(National Centers for Environmental Prediction) / NCAR(National Centers for Atmospheric Research) reanalysis data. The heavy rainfall is accompanied with mesoscale rain clouds developing over the Southwest region of Korea during the period from 0300 LST to 2100 LST 25 June 2006. The surface cyclone in the Changma front is generated and developed rapidly when it meets following vertical conditions: The maximum value of relative vorticity is appeared at 700 hPa and is extended gradually near the surface. It is thought that the vertical structure of relative vorticity is closely related with the descent of strong wind zone exceeding $10ms^{-1}$. The jet core at 200 hPa is shifted southward and extended downward and the low-level jet stream associated with upper-level jet stream appeared at 850 hPa. Kinematic features of heavy rainfall system at cyclone-generating point are as follows: In the generating stage of cyclone, the relative vorticity below 850 hPa increased and the convergence below 850 hPa and the divergence at 400 hPa are intensified by southward movement of jet core at 200 hPa. The heavy rainfall system seems to locate to the south of the exit region of upper-level jet streak; In the developing stage of cyclone, the relative vorticity below 850 hPa and the convergence near surface are further strengthened and upward vertical velocity between 850 hPa and 200 hPa is increased.

• Korean Institute of Interior Design Journal
• /
• v.26 no.3
• /
• pp.45-54
• /
• 2017

A social phenomenon that occurs in a physical space is said to be a complex system. However, space syntax, which is commonly employed by researchers to identify such social phenomena, has various limitations in interpreting their complexity. On the other hand, agent-based modeling considers a variety of factors including the personality of the agent, objective-oriented work flows, estimation according to time flows and better prediction of space use through diverse parameters depending the situation, as well as the characteristics of the space. The agent-based method thus has the potentials to be developed as an alternative to space syntax techniques. In particular, discrete event driven simulation(DEVS), which is part of the agent-based modeling method, embraces the concept of networks just like space syntax, which allows a possible theoretical linkage in the future. This study suggests a procedural model of agent-based DEVS reflecting two different connection methods, i.e. connections between adjacent areas and those of the entire space, and attempts to identify the relationship between the local and regional indices of space syntax. A number of spaces were selected as examples-one for a preliminary experiment and eight modified for the main experiment-and space syntax and DEVS were applied to each of them. The comparative analysis of the results led to the conclusions as follows: 1) Adjacent connections were closely related to local indices, while the whole-space approach to regional indices. Local integration shows both characteristics. 2) Observation of the time flow model indicated a faster convergence with the range of 1 to 3-fold of the total time of one lap, with the error of less than 10%. 3) The heat map analysis showed more obvious characteristics of using the space for the entire space rather than adjacent connections. 4) Space syntax shows higher eligibility than ABM.

• Atmosphere
• /
• v.26 no.2
• /
• pp.277-288
• /
• 2016

Polar lows are intense mesoscale cyclones that mainly occur over the sea in polar regions. Owing to their small spatial scale of a diameter less than 1000 km, simulating polar lows is a challenging task. At King Sejong station in West Antartica, polar lows are often observed. Despite the recent significant climatic changes observed over West Antarctica, adequate validation of regional simulations of extreme weather events such as polar lows are rare for this region. To address this gap, simulation results from a recent version of the Polar Weather Research and Forecasting model (Polar WRF) covering Antartic Peninsula at a high horizontal resolution of 3 km are validated against near-surface meteorological observations. We selected a case of high wind speed event on 7 January 2013 recorded at Automatic Meteorological Observation Station (AMOS) in King Sejong station, Antarctica. It is revealed by in situ observations, numerical weather prediction, and reanalysis fields that the synoptic and mesoscale environment of the strong wind event was due to the passage of a strong mesoscale polar low of center pressure 950 hPa. Verifying model results from 3 km grid resolution simulation against AMOS observation showed that high skill in simulating wind speed and surface pressure with a bias of $-1.1m\;s^{-1}$ and -1.2 hPa, respectively. Our evaluation suggests that the Polar WRF can be used as a useful dynamic downscaling tool for the simulation of Antartic weather systems and the near-surface meteorological instruments installed in King Sejong station can provide invaluable data for polar low studies over West Antartica.

• Nuclear Engineering and Technology
• /
• v.40 no.7
• /
• pp.615-624
• /
• 2008

In this paper, the contribution of task types and error types involved in the human-related unplanned reactor trip events that have occurred between 1986 and 2006 in Korean nuclear power plants are analysed in order to establish a strategy for reducing the human-related unplanned reactor trips. Classification systems for the task types, error modes, and cognitive functions are developed or adopted from the currently available taxonomies, and the relevant information is extracted from the event reports or judged on the basis of an event description. According to the analyses from this study, the contributions of the task types are as follows: corrective maintenance (25.7%), planned maintenance (22.8%), planned operation (19.8%), periodic preventive maintenance (14.9%), response to a transient (9.9%), and design/manufacturing/installation (6.9%). According to the analysis of the error modes, error modes such as control failure (22.2%), wrong object (18.5%), omission (14.8%), wrong action (11.1 %), and inadequate (8.3%) take up about 75% of the total unplanned trip events. The analysis of the cognitive functions involved in the events indicated that the planning function had the highest contribution (46.7%) to the human actions leading to unplanned reactor trips. This analysis concludes that in order to significantly reduce human-induced or human-related unplanned reactor trips, an aide system (in support of maintenance personnel) for evaluating possible (negative) impacts of planned actions or erroneous actions as well as an appropriate human error prediction technique, should be developed.

• Journal of Astronomy and Space Sciences
• /
• v.34 no.4
• /
• pp.315-330
• /
• 2017

Halo coronal mass ejections (CMEs) originating from solar activities give rise to geomagnetic storms when they reach the Earth. Variations in the geomagnetic field during a geomagnetic storm can damage satellites, communication systems, electrical power grids, and power systems, and induce currents. Therefore, automated techniques for detecting and analyzing halo CMEs have been eliciting increasing attention for the monitoring and prediction of the space weather environment. In this study, we developed an algorithm to sense and detect halo CMEs using large angle and spectrometric coronagraph (LASCO) C3 coronagraph images from the solar and heliospheric observatory (SOHO) satellite. In addition, we developed an image processing technique to derive the morphological and dynamical characteristics of halo CMEs, namely, the source location, width, actual CME speed, and arrival time at a 21.5 solar radius. The proposed halo CME automatic analysis model was validated using a model of the past three halo CME events. As a result, a solar event that occurred at 03:38 UT on Mar. 23, 2014 was predicted to arrive at Earth at 23:00 UT on Mar. 25, whereas the actual arrival time was at 04:30 UT on Mar. 26, which is a difference of 5 hr and 30 min. In addition, a solar event that occurred at 12:55 UT on Apr. 18, 2014 was estimated to arrive at Earth at 16:00 UT on Apr. 20, which is 4 hr ahead of the actual arrival time of 20:00 UT on the same day. However, the estimation error was reduced significantly compared to the ENLIL model. As a further study, the model will be applied to many more events for validation and testing, and after such tests are completed, on-line service will be provided at the Korean Space Weather Center to detect halo CMEs and derive the model parameters.