• Journal of Digital Contents Society
• /
• v.19 no.4
• /
• pp.789-799
• /
• 2018

This article presents machine learning based approach on Big data to analyzing time series data for anomaly detection in such industrial complex system. Long Short-Term Memory (LSTM) network have been demonstrated to be improved version of RNN and have become a useful aid for many tasks. This LSTM based model learn the higher level temporal features as well as temporal pattern, then such predictor is used to prediction stage to estimate future data. The prediction error is the difference between predicted output made by predictor and actual in-coming values. An error-distribution estimation model is built using a Gaussian distribution to calculate the anomaly in the score of the observation. In this manner, we move from the concept of a single anomaly to the idea of the collective anomaly. This work can assist the monitoring and management of Smart Factory in minimizing failure and improving manufacturing quality.

• Water for future
• /
• v.27 no.3
• /
• pp.35-43
• /
• 1994

Because a failure to provide enough plunge pool depth can create a risk to the structural stability of the psillways or dams, many researchers have proposed experimental formulas for claculating ultimate scour depth under jet issued from spillways and pipe culverts. For the design purposes of a plunge pool, scour potential distribution is important as much as the ultimate scour depth is. In this study scour potential distributions near the jet impinging point on a porous plane which can simulate a real cohesionless movable flat bed has been measured. Experimental results showed that scour potential distributions are geometrically similar to each other provided the angle of jet impact was the same. Statistical analysis of experimental results showed that scour potential distributions for the design purposes of a plunge pool could be expressed by a single equation within a range of this experiment. The proposed formula for the prediction of scour potential distributions agrees well with experimental measurements.

• Journal of Korean Society on Water Environment
• /
• v.23 no.3
• /
• pp.356-366
• /
• 2007

In this paper a methodology is developed to prioritize replacement of water distribution pipes according to the economical efficiency of replacement and assess the long-term effects of water main replacement policies on water distribution systems. The methodology is implemented with MATLAB to develop a computer algorithm which is used to apply the methodology to a case study water distribution system. A pipe break prediction model is used to estimate future costs of pipe repair and replacement, and the economically optimal replacement time of a pipe is estimated by obtaining the time at which the present worth of the total costs of repair and replacement is minimum. The equation for estimating the present worth of the total cost is modified to reflect the fact that a pipe can be replaced in between of failure events. The results of the analyses show that about 9.5% of the pipes in the case study system is required to be replaced within the planning horizon. Analyses of the yearly pipe replacement requirements for the case study system are provided along with the compositions of the replacement. The effects of water main replacement policies, for which yearly replacement length scenario and yearly replacement budget scenario are used, during a planning horizon are simulated in terms of the predicted number of pipe failures and the saved repair costs.

• Smart Structures and Systems
• /
• v.25 no.6
• /
• pp.733-749
• /
• 2020

This study presents a computer vision-based approach for representing time evolution of structural damages leveraging a database of inspection images. Spatially incoherent but temporally sorted archival images captured by robotic cameras are exploited to represent the damage evolution over a long period of time. An access to a sequence of time-stamped inspection data recording the damage growth dynamics is premised to this end. Identification of a structural defect in the most recent inspection data set triggers an exhaustive search into the images collected during the previous inspections looking for correspondences based on spatial proximity. This is followed by a view synthesis from multiple candidate images resulting in a single reconstruction for each inspection round. Cracks on concrete surface are used as a case study to demonstrate the feasibility of this approach. Once the chronology is established, the damage severity is quantified at various levels of time scale documenting its progression through time. The proposed scheme enables the prediction of damage severity at a future point in time providing a scope for preemptive measures against imminent structural failure. On the whole, it is believed that the present study will immensely benefit the structural inspectors by introducing the time dimension into the autonomous condition assessment pipeline.

• Environmental Engineering Research
• /
• v.18 no.1
• /
• pp.37-43
• /
• 2013

Water distribution pipes installed underground have potential risks of pipe failure and burst. After years of use, pipe walls tend to be corroded due to aggressive soil environments where they are located. The present study aims to assess the degree of external corrosion of a distribution pipe network. In situ data obtained through test pit excavation and direct sampling are carefully collated and assessed. A statistical approach is useful to predict severity of pipe corrosion at present and in future. First, criteria functions defined by discriminant function analysis are formulated to judge whether the pipes are seriously corroded. Data utilized in the analyses are those related to soil property, i.e., soil resistivity, pH, water content, and chloride ion. Secondly, corrosion factors that significantly affect pipe wall pitting (vertical) and spread (horizontal) on the pipe surface are identified with a view to quantifying a degree of the pipe corrosion. Finally, a most reliable model represented in the form of a multiple regression equation is developed for this purpose. From these analyses, it can be concluded that our proposed model is effective to predict the severity and rate of pipe corrosion utilizing selected factors that reflect the fuzzy soil environment.

• Journal of the Korean Society of Safety
• /
• v.38 no.1
• /
• pp.87-92
• /
• 2023

The number of unmanned light rail train operators is continuously increasing in Korea. In a failure event during an operation due to the nature of the unmanned operation, recovery is performed based on the remote control. However, if remote recovery is not feasible, safety personnel arrive at the train to resume the train operation. There are regulations on safety personnel and the suspension time of the train operation. However, there is currently no rule for safety personnel deployment. Currently, railway operating organizations operate in three scenarios: safety personnel on board trains, stationed at stations, and deployed at major stations. Four major factors influence the downtime for each emergency response scenario. However, these four influencing factors vary too much to predict results with simple calculations. In this study, four influencing factors were considered as random variables with high uncertainty. In addition, the Monte Carlo method was applied to each scenario for the safety personnel deployment to predict train service downtime. This study found a 17% difference in train service suspension by safety personnel deployment scenario. The results of this study can be used in setting service goals, such as standards for future safety personnel placement and frequency of service interruptions.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2019.05a
• /
• pp.230-231
• /
• 2019

This study is a development on condition based maintenance(CBM) technology which is a core item of future autonomous ships. It is developing to design & installation of condition monitoring system and acquisition & processing of data from ongoing ships for fault prediction & prognosis of engine in operation. The ultimate goal of this study is to develop a predicts and decision support software for marine engine faults. To do this, the FMEA and fault tree analysis of the main engine should be accompanied by the analysis of classification of system, identification of the components, the type of faults, and the cause and phenomenon of the failure. Finally, the CBM system solution software could predict and diagnose the failure of main engine through integrated analysis for bid-data of ongoing ships and engineering knowledge. Through this study, it is possible to pro-actively cope with abnormal signals of engine and to manage efficiently, and as a result, expected that marine accident and ship operation loss during navigation will be prevented in advance.

• Journal of Intelligence and Information Systems
• /
• v.24 no.4
• /
• pp.1-32
• /
• 2018

In addition to stakeholders including managers, employees, creditors, and investors of bankrupt companies, corporate defaults have a ripple effect on the local and national economy. Before the Asian financial crisis, the Korean government only analyzed SMEs and tried to improve the forecasting power of a default prediction model, rather than developing various corporate default models. As a result, even large corporations called 'chaebol enterprises' become bankrupt. Even after that, the analysis of past corporate defaults has been focused on specific variables, and when the government restructured immediately after the global financial crisis, they only focused on certain main variables such as 'debt ratio'. A multifaceted study of corporate default prediction models is essential to ensure diverse interests, to avoid situations like the 'Lehman Brothers Case' of the global financial crisis, to avoid total collapse in a single moment. The key variables used in corporate defaults vary over time. This is confirmed by Beaver (1967, 1968) and Altman's (1968) analysis that Deakins'(1972) study shows that the major factors affecting corporate failure have changed. In Grice's (2001) study, the importance of predictive variables was also found through Zmijewski's (1984) and Ohlson's (1980) models. However, the studies that have been carried out in the past use static models. Most of them do not consider the changes that occur in the course of time. Therefore, in order to construct consistent prediction models, it is necessary to compensate the time-dependent bias by means of a time series analysis algorithm reflecting dynamic change. Based on the global financial crisis, which has had a significant impact on Korea, this study is conducted using 10 years of annual corporate data from 2000 to 2009. Data are divided into training data, validation data, and test data respectively, and are divided into 7, 2, and 1 years respectively. In order to construct a consistent bankruptcy model in the flow of time change, we first train a time series deep learning algorithm model using the data before the financial crisis (2000~2006). The parameter tuning of the existing model and the deep learning time series algorithm is conducted with validation data including the financial crisis period (2007~2008). As a result, we construct a model that shows similar pattern to the results of the learning data and shows excellent prediction power. After that, each bankruptcy prediction model is restructured by integrating the learning data and validation data again (2000 ~ 2008), applying the optimal parameters as in the previous validation. Finally, each corporate default prediction model is evaluated and compared using test data (2009) based on the trained models over nine years. Then, the usefulness of the corporate default prediction model based on the deep learning time series algorithm is proved. In addition, by adding the Lasso regression analysis to the existing methods (multiple discriminant analysis, logit model) which select the variables, it is proved that the deep learning time series algorithm model based on the three bundles of variables is useful for robust corporate default prediction. The definition of bankruptcy used is the same as that of Lee (2015). Independent variables include financial information such as financial ratios used in previous studies. Multivariate discriminant analysis, logit model, and Lasso regression model are used to select the optimal variable group. The influence of the Multivariate discriminant analysis model proposed by Altman (1968), the Logit model proposed by Ohlson (1980), the non-time series machine learning algorithms, and the deep learning time series algorithms are compared. In the case of corporate data, there are limitations of 'nonlinear variables', 'multi-collinearity' of variables, and 'lack of data'. While the logit model is nonlinear, the Lasso regression model solves the multi-collinearity problem, and the deep learning time series algorithm using the variable data generation method complements the lack of data. Big Data Technology, a leading technology in the future, is moving from simple human analysis, to automated AI analysis, and finally towards future intertwined AI applications. Although the study of the corporate default prediction model using the time series algorithm is still in its early stages, deep learning algorithm is much faster than regression analysis at corporate default prediction modeling. Also, it is more effective on prediction power. Through the Fourth Industrial Revolution, the current government and other overseas governments are working hard to integrate the system in everyday life of their nation and society. Yet the field of deep learning time series research for the financial industry is still insufficient. This is an initial study on deep learning time series algorithm analysis of corporate defaults. Therefore it is hoped that it will be used as a comparative analysis data for non-specialists who start a study combining financial data and deep learning time series algorithm.

• Journal of the Korea Concrete Institute
• /
• v.22 no.6
• /
• pp.777-786
• /
• 2010

This paper examines the flexural behavior of full-scale prestressed concrete girders that were constructed of steel fiber reinforced ultra high performance concrete (UHPC). This study is designed to provide more information about the bending characteristics of UHPC girders in order to establish a reasonable prediction model for flexural resistance and deflection for future structural design codes. Short steel fibers have been introduced into prestressed concrete T-girders in order to study their effects under flexural loads. Round straight high strength steel fibers were used at volume fraction of 2%. The girders were cast using 150~190 MPa steel fiber reinforced UHPC and were designed to assess the ability of steel fiber reinforced UHPC to carry flexural loads in prestressed girders. The experimental results show that steel fiber reinforced UHPC enhances the cracking behavior and ductility of beams. Moreover, when ultimate failure did occur, the failure of girders composed of steel fiber reinforced UHPC was observed to be precipitated by the pullout of steel fibers that were bridging tension cracks in the concrete. Flexural failure of girders occurred when the UHPC at a particular cross section began to lose tensile capacity due to steel fiber pullout. In addition, it was determined that the level of prestressing force influenced the ultimate load capacity.

• Journal of Korea Water Resources Association
• /
• v.53 no.9
• /
• pp.661-670
• /
• 2020

The water supply system has a wider installation range and various components of it than other infrastructure, making it difficult to secure stability against earthquakes. Therefore, it is necessary to develop methods for evaluating the seismic performance of water supply systems. Ground Motion Prediction Equation (GMPE) is used to evaluate the seismic performance (e.g, failure probability) for water supply facilities such as pump, water tank, and pipes. GMPE is calculated considering the independent variables such as the magnitude of the earthquake and the ground motion such as PGV (Peak Ground Velocity) and PGA (Peak Ground Acceleration). Since the large magnitude earthquake data has not accumulated much to date in Korea, this study tried to select a suitable GMPE for the domestic earthquake simulation by using the earthquake data measured in Korea. To this end, GMPE formula is calculated based on the existing domestic earthquake and presented the results. In the future, it is expected that the evaluation will be more appropriate if the determined GMPE is used when evaluating the seismic performance of domestic waterworks. Appropriate GMPE can be directly used to evaluate hydraulic seismic performance of water supply networks. In other words, it is possible to quantify the damage rate of a pipeline during an earthquake through linkage with the pipe failure probability model, and it is possible to derive more reasonable results when estimating the water outage or low-pressure area due to pipe damages. Finally, the quantifying result of the seismic performance can be used as a design criteria for preparing an optimal restoration plan and proactive seismic design of pipe networks to minimize the damage in the event of an earthquake.