• Journal of Advanced Navigation Technology
• /
• v.23 no.5
• /
• pp.380-385
• /
• 2019

This paper is concerned with departure passengers' dwell time analysis using real system data. Previous researches emphasize the importance of dwell time analysis from perspective of airport terminal planning and non-aeronautical revenue. However, short-term airport operation using passengers' dwell time is considered impossible due to absence of passengers' behavior data. Recently, in accordance with the wave of smart airport, world leading airports are systematically collecting passenger data. So there is high possibility of analyzing passengers' dwell time with the data stacked in the airport database. We conducted dwell time analysis using data from Incheon Int'l airport. In order to handle passenger data, we adapted clustering algorithm which is one of data mining techniques. As a clustering result, passengers are divided into 3 clusters. One is the cluster for passengers whose dwell time is relatively short and who tend to spend longer time in the airside. Another is the cluster for passengers who have near 3 hours dwell time. The other is the cluster for passengers whose total dwell time is extremely long.

• Journal of the Society of Naval Architects of Korea
• /
• v.32 no.4
• /
• pp.38-47
• /
• 1995

The three-dimensional multi-layer computer simulation technology for tidal current developed in the previous study is updated to a new version. many improvements are achieved by following changes : (1) No-reflection condition is adopted instead of no-gradient condition as an open boundary condition. (2) Time marching algorithm is changed so that velocity and pressure(surface movement) might be salved in turn at different time step (3) Convection term in equation of motion is estimated by upwind differencing scheme instead of central differencing. The stability is improved considerably and the steady state is achieved within 2 tidal periods which is about 3 times shorter than that of the old version. Moreover, fluctuations in time disappeared by introducing the new time marching technique. An application to the real near shore area(near Inchon harbor) is performed by the new version. Simulated results are compared with those by the simulation total developed in the University of Tokyo. Validity and effectiveness of the two simulation technologies are chocked through the comparative research works.

• Journal of the Society of Naval Architects of Korea
• /
• v.54 no.4
• /
• pp.329-334
• /
• 2017

As the motion response of heave for floating bodies on the water surface is relatively large near the natural frequency, it is necessary to predict its value accurately from the stage of initial design. Bodies accelerating in fluid experience force acted upon by the fluid, and this force is quantified by using the concept of added mass. For predicting the natural frequency of heave we need to know the added mass, which is given as a function of frequency, and hence the natural frequency can be obtained through only by iteration process, as was pointed out by Lee (2008). His method was applied to circular cylinders, and two dimensional cylinders of Lewis form by making use of the Ursell-Tasai method in the previous works, Lee and Lee (2013), Kim and Lee (2013), and Song and Lee (2015). In this work, a similar algorithm employing the concept of strip method is adopted for predicting the heave natural frequency of KCS(KRISO Container Ship), and the obtained computational result was compared with other existing experimental data, and the agreement seems reasonable. Furthermore, through the error analysis, it is shown that why the frequency corresponding to the local minimum of the added mass and the natural frequency are very close. And it seems probable that we can predict the heave natural frequency if we know only the local minimum of added mass and the corresponding frequency under a condition, which holds for ship-like bodies in general.

• Transactions of Materials Processing
• /
• v.31 no.6
• /
• pp.351-364
• /
• 2022

The yield criterion, or called yield function, plays an important role in the study of plastic working of a sheet because it governs the plastic deformation properties of the sheet during plastic forming process. In this paper, we propose a modified version of previous anisotropic yield function (Trans. Mater. Process., 31(4) 2022, pp. 214-228) based on J2 and J3 stress invariants. The proposed anisotropic yield model has the 6th-order of stress components. The modified version of the anisotropic yield function in this study is as follows. f(J₂⁰,J₃⁰) ≡ (J₂⁰)³ + α(J₃⁰)² + β(J₂⁰)^3/2 × (J₃⁰) = k⁶ The proposed anisotropic yield function well explains the anisotropic plastic behavior of various sheets such as aluminum, high strength steel, magnesium alloy sheets etc. by introducing the parameters α and β, and also exhibits both symmetrical and asymmetrical yield surfaces. The parameters included in the proposed model are determined through an optimization algorithm from uniaxial and biaxial experimental data under proportional loading path. In this study, the validity of the proposed anisotropic yield function was verified by comparing the yield surface shape, normalized uniaxial yield stress value, and Lankford's anisotropic coefficient R-value derived with the experimental results. Application for the proposed anisotropic yield function to AA6016-T4 aluminum and DP980 sheets shows symmetrical yielding behavior and to AZ31B magnesium shows asymmetric yielding behavior, it was shown that the yield locus and yielding behavior of various types of sheet materials can be predicted reasonably by using the proposed anisotropic yield function.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.8
• /
• pp.4123-4141
• /
• 2019

Autonomous driving technology is divided into 0~5 levels. Of these, Level 5 is a fully autonomous vehicle that does not require a person to drive at all. The automobile industry has been trying to develop Level 5 to satisfy safety, but commercialization has not yet been achieved. In order to commercialize autonomous unmanned vehicles, there are several problems to be solved for driving safety. To solve one of these, this paper proposes 'A Deep Learning Part-diagnosis Platform(DLPP) based on an In-vehicle On-board gateway for an Autonomous Vehicle' that diagnoses not only the parts of a vehicle and the sensors belonging to the parts, but also the influence upon other parts when a certain fault happens. The DLPP consists of an In-vehicle On-board gateway(IOG) and a Part Self-diagnosis Module(PSM). Though an existing vehicle gateway was used for the translation of messages happening in a vehicle, the IOG not only has the translation function of an existing gateway but also judges whether a fault happened in a sensor or parts by using a Loopback. The payloads which are used to judge a sensor as normal in the IOG is transferred to the PSM for self-diagnosis. The Part Self-diagnosis Module(PSM) diagnoses parts itself by using the payloads transferred from the IOG. Because the PSM is designed based on an LSTM algorithm, it diagnoses a vehicle's fault by considering the correlation between previous diagnosis result and current measured parts data.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.1
• /
• pp.27-32
• /
• 2019

Research that implements the classification process through Deep Learning algorithm, one of the outstanding human abilities, includes a unimodal model, a multi-modal model, and a multi-modal method using music videos. In this study, the results were better by suggesting a system to analyze each song's spectrum into short samples and vote for the results. Among Deep Learning algorithms, CNN showed superior performance in the category of music genre compared to RNN, and improved performance when CNN and RNN were applied together. The system of voting for each CNN result by Deep Learning a short sample of music showed better results than the previous model and the model with Softmax layer added to the model performed best. The need for the explosive growth of digital media and the automatic classification of music genres in numerous streaming services is increasing. Future research will need to reduce the proportion of undifferentiated songs and develop algorithms for the last category classification of undivided songs.

• Journal of the Korea Computer Graphics Society
• /
• v.24 no.5
• /
• pp.21-30
• /
• 2018

In this paper, we use Sound-tracing, a 3D sound technology based on ray-tracing that uses geometric method as auditory technology to enhance realism. The Sound-tracing is costly in the sound propagation stage. In order to reduce the sound propagation cost, we propose a method to calculate the average effective frame number of previous frames using the frame coherence property and to adjust the depth according to the space based on the calculated number. Experimental results show that the path loss rate is 0.72% and the traversal & Intersection test calculation amount is decreased by 85.13% and the frame rate is increased by 4.48% when the sound source is indoors, compared with the result of the case without depth control. When the sound source was outdoors, the path loss was 0% and the traversal & Intersection test calculation amount is decreased by 25.01% and the frame rate increased by 7.85%. This allowed the rendering performance to be increased while minimizing the path loss rate.

• The Journal of Information Systems
• /
• v.28 no.3
• /
• pp.249-276
• /
• 2019

Purpose The purpose of this study is to develop a prediction model and decision rules for the elderly's suicidal ideation based on the Korean Welfare Panel survey data. By utilizing this data, we obtained many decision rules to predict the elderly's suicide ideation. Design/methodology/approach This study used classification analysis to derive decision rules to predict on the basis of decision tree technique. Weka 3.8 is used as the data mining tool in this study. The decision tree algorithm uses J48, also known as C4.5. In addition, 66.6% of the total data was divided into learning data and verification data. We considered all possible variables based on previous studies in predicting suicidal ideation of the elderly. Finally, 99 variables including the target variable were used. Classification analysis was performed by introducing sampling technique through backward elimination and data balancing. Findings As a result, there were significant differences between the data sets. The selected data sets have different, various decision tree and several rules. Based on the decision tree method, we derived the rules for suicide prevention. The decision tree derives not only the rules for the suicidal ideation of the depressed group, but also the rules for the suicidal ideation of the non-depressed group. In addition, in developing the predictive model, the problem of over-fitting due to the data imbalance phenomenon was directly identified through the application of data balancing. We could conclude that it is necessary to balance the data on the target variables in order to perform the correct classification analysis without over-fitting. In addition, although data balancing is applied, it is shown that performance is not inferior in prediction rate when compared with a biased prediction model.

• Journal of Internet Computing and Services
• /
• v.20 no.1
• /
• pp.57-65
• /
• 2019

In the case of Domeggook B2B online shopping malls, it has a market share of over 70% with more than 2 million members and 800,000 items are sold per one day. However, since the same or similar items are stored and registered in different catalog entries, it is difficult for the buyer to search for items, and problems are also encountered in managing B2B large shopping malls. Therefore, in this study, we developed a catalog entry auto classification and recommendation system for products by using semi-supervised machine learning method based on previous huge shopping mall purchase information. Specifically, when the seller enters the item registration information in the form of natural language, KoNLPy morphological analysis process is performed, and the Naïve Bayes classification method is applied to implement a system that automatically recommends the most suitable catalog information for the article. As a result, it was possible to improve both the search speed and total sales of shopping mall by building accuracy in catalog entry efficiently.

• The Journal of the Acoustical Society of Korea
• /
• v.38 no.1
• /
• pp.23-29
• /
• 2019

In this paper, a deep learning-based detection and classification using multi-band frequency signals is presented for detecting earthquakes prevalent in Korea. Based on an analysis of the previous earthquakes in Korea, it is observed that multi-band signals are appropriate for classifying earthquake signals. Therefore, in this paper, we propose a deep CNN (Convolutional Neural Network) using multi-band signals as training data. The proposed algorithm extracts the multi-band signals (Low/Medium/High frequency) by applying band pass filters to mel-spectrum of earthquake signals. Then, we construct three CNN architecture pipelines for extracting features and classifying the earthquake signals by a late fusion of the three CNNs. We validate effectiveness of the proposed method by performing various experiments for classifying the domestic earthquake signals detected in 2018.