• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.05a
• /
• pp.540-543
• /
• 2020

본 논문에서는 대표적인 이미지 분류 모델인 CNN(Convolutional Neural Network)과 시간에 따른 이미지의 변화를 학습할 수 있는 LSTM(Long Short-Term Memory) 기반의 온라인 서명인식 모델을 제안한다. 실제로는 위조서명을 미리 구하기 어렵다는 사실을 고려해 서명검증 대상자가 아닌 타인의 진서명과 대상자의 일반 필기 데이터를 음의 데이터로서 학습에 사용하였다. 실험 결과, 전체 이미지 중 서명 부분의 비율에 따라 좋은 성능을 보이는 검증 모델이 다르며 Accuracy 성능지표를 통해 이 비율이 높거나 낮을 경우 CNN-LSTM 이, 중간일 경우 CNN 이 적합하다는 것을 확인하였다.

• Smart Structures and Systems
• /
• v.11 no.6
• /
• pp.637-662
• /
• 2013

The Jiangyin Bridge is a suspension bridge with a main span of 1385 m over the Yangtze River in Jiangsu Province, China. Being the first bridge with a main span exceeding 1 km in Chinese mainland, it had been instrumented with a structural health monitoring (SHM) system when completed in 1999. After operation for several years, it was found with malfunction in sensors and data acquisition units, and insufficient sensors to provide necessary information for structural health evaluation. This study reports the SHM system upgrade project on the Jiangyin Bridge. Although implementations of SHM system have been reported worldwide, few studies are available on the upgrade of SHM system so far. Recognizing this, the upgrade of original SHM system for the bridge is first discussed in detail. Especially, lessons learned from the original SHM system are applied to the design of upgraded SHM system right away. Then, performance assessment of the bridge, including: (i) characterization of temperature profiles and effects; (ii) recognition of wind characteristics and effects; and (iii) identification of modal properties, is carried out by making use of the long-term monitoring data obtained from the upgraded SHM system. Emphasis is placed on the verification of design assumptions and prediction of bridge behavior or extreme responses. The results may provide the baseline for structural health evaluation.

• Geomechanics and Engineering
• /
• v.20 no.6
• /
• pp.517-525
• /
• 2020

The long-term stability of rock engineering is significantly affected by the time-dependent deformation behavior of rock, which is an important mechanical property of rock for engineering design. Although the hard rocks show small creep deformation, it cannot be ignored under high-stress condition during deep excavation. The inner mechanism of creep is complicated, therefore, it is necessary to investigate the relationship between microscopic creep mechanism and the macro creep behavior of rock. Microscopic numerical modeling of sandstone creep was performed in the investigation. A numerical sandstone sample was generated and Parallel Bond contact and Burger's contact model were assigned to the contacts between particles in DEM simulation. Sensitivity analysis of the microscopic creep parameters was conducted to explore how microscopic parameters affect the macroscopic creep deformation. The results show that the microscopic creep parameters have linear correlations with the corresponding macroscopic creep parameters, whereas the friction coefficient shows power function with peak strength and Young's modulus, respectively. Moreover, the microscopic parameters were calibrated. The creep modeling curve is in good agreement with the verification test result. Finally, the creep curves under one-step loading and multi-step loading were compared. This investigation can act as a helpful reference for modeling rock creep behavior from a microscopic mechanism perspective.

• The Korean Journal of Elementary Counseling
• /
• v.6 no.1
• /
• pp.49-71
• /
• 2007

The purpose of this study was to examine the effects of a career group counseling program on the career maturity and achievement motivation of elementary school underachievers. This study targeted the fifth grader underachievers. Twenty underachievers, whose pre-test homogeneity in career maturity and achievement motivation was verified, were allocated to an experimental group and a control group by 10 underachievers each. The program was conducted for a total of 12 sessions, twice per week, and each session lasted 80 minutes. For verification of the program effects in this study, MANOVA and ANOVA were carried out for pre- and post-score tests between the groups, and repeated-MANOVA and paired t-test were peformed to look into the changes within a group by using SPSS/WIN(version 12.0) in terms of statistical analyses. The findings in this study are as follows: First, as a result of post test for the experimental and control groups, and difference verification of post and pre-tests within the experimental group, both career maturity and sub-factors of career maturity demonstrated significant differences. Second, as a result of difference verification of the pre- and post-tests within the experimental group, achievement motivation significantly improved, but there was no statistically significant difference in achievement motivation in the post-test of experimental and control groups. However, of the sub-factors of the achievement motivation, self-confidence and Interest in the long-term future significantly improved. Third, we can see that the career group counseling program provided opportunities for the underachievers to search their own character, interest, and aptitude, and broadly encounter the world of occupations. Also, we can infer that the underachieves had active and developmental attitudes towards future dreams and their current efforts. As noted above, the career group counseling program proved that it has a positive effect to the improvement of the elementary school underachievers' career maturity and achievement motivation in some variables.

• Proceedings of the KIEE Conference
• /
• 2009.07a
• /
• pp.294_295
• /
• 2009

Superconducting fault current limiter (SFCL) is an power device of a novel concept. While SFCLs generate no ohmic loss during the operation carrying normal currents, they can limit fault currents very fast making large impedance by their quench characteristics. In 2006, KEPCO has developed a distribution class hybrid type SFCL by a collaborative research project with LS industrial systems. The SFCL has merits in practical and economical points of view. In the SFCL, the superconductor just plays a role of a fault detector and the current limiting is completed by the other current limiting element made of normal metals throu호 the line commutation. As a result, the required amounts of superconductors can be reduced considerably. Consequently, the hybrid SFCL can be fabricated with small size and cost, maintaining perfect current limiting performance. Currently, KEPCO is carrying out a research project at Gochang power test center for the purpose of the verification test of the 22.9 kV/ 630 A class SFCL for the practical application in real grid. Through the project, a long term operational test and fault current test will be done. In this paper, the back ground of development and installation of the SFCL will be explained and the operation plan of the SFCL for the verification test is also introduced.

• The Journal of the Acoustical Society of Korea
• /
• v.19 no.5
• /
• pp.20-26
• /
• 2000

In speaker verification system the HMM(hidden Markov model) parameter updating using small amount of data and the priori threshold decision are crucial factor for dealing with long-term variability in people voices. In the paper we present the speaker model updating technique which can be adaptable to the session-to-intra speaker variability and the priori threshold determining technique. The proposed technique decreases verification error rates which the session-to-session intra-speaker variability can bring by adapting new speech data to speaker model parameter through Baum Welch re-estimation. And in this study the proposed priori threshold determining technique is decided by a hybrid score measurement which combines the world model based technique and the cohen model based technique together. The results show that the proposed technique can lead a better performance and the difference of performance is small between the posteriori threshold decision based approach and the proposed priori threshold decision based approach.

• Journal of The Korean Society of Integrative Medicine
• /
• v.6 no.3
• /
• pp.43-57
• /
• 2018

Purpose : The act of reporting child abuse by people obliged to report child abuse is a way to detect long-term incidents of child abuse in early stages and to prevent a vicious cycle of social crimes. Thus, this study intends to enhance the reporting of child abuse through a study on the perception of people obliged to report child abuse through the public service announcement on child abuse. Methods : First, in order to identify the demographic characteristics of the subjects, the frequency and percentage were calculated. Second, in order to verify the reliability of the variables, Cronbach ${\alpha}$ was calculated. Third, in order to examine the perception of people obliged to report child abuse and the perception of reporting, a t-test (verification) and One-way ANOVA were conducted. Fourth, in order to identify the relationship between variables, a Correlation analysis was conducted. Result : The results show that people obliged to report child abuse did not perceive the public service announcement on child abuse positively and it varied depending on their jobs. The correlation between the public service announcement on child abuse and sub-factor of perceptions of reporting showed that a positive correlation was found in reliability, interest, understanding, perceiving the public service announcement on child abuse, perceiving severity and effectiveness, whereas a negative correlation was found in perceiving hazards. Conclusion : The education for preventing child abuse is required by the job type of people obliged to report. It is necessary to convey the message of effective public service announcement. The public service announcements on child abuse must be nationwide and long-term projects.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1328-1336
• /
• 2019

To maintain the safe and optimal performance of batteries, accurate estimation of state of charge (SOC) is critical. In this paper, Long short-term memory network (LSTM) based on the artificial intelligence algorithm is applied to address the problem of the conventional coulomb-counting method. Different discharge cycles are concatenated to form the dataset for training and verification. In oder to improve the quality of input data for learning, preprocessing was performed. In addition, we compared learning ability and SOC estimation performance according to the structure of LSTM model and hyperparameter setup. The trained model was verified with a UDDS profile and achieved estimated accuracy of RMSE 0.82% and MAX 2.54%.

• Journal of the Society of Naval Architects of Korea
• /
• v.53 no.5
• /
• pp.380-387
• /
• 2016

Present research target to develop the procedure of long-term fatigue analysis of the structural details near the upper rolling chock of IMO type B tank by using the time domain modal analysis technique where both the contact and friction behavior can be accurately simulated. In order to perform the time domain analysis focused on the contact and friction, the entire model of the hull and tank was condensed with DOF reduction technique, which is obtained by transforming the global finite element model into its quasi-static modal coordinate. Modal analysis using the quasi-static deformation modes is chosen as a cost effective time domain simulation method and this is based on the fact that the structural response of the tank is quasi-static. Based on the developed cost effective time domain simulation method, the long-term fatigue analysis procedure for the structural details near the rolling chock and key of independent type tank is targeted to be established. The developed fatigue assessment procedure takes into account, wave induced stress and both contact and friction induced stress without loss of accuracy.

• Journal of the Korea Safety Management & Science
• /
• v.25 no.4
• /
• pp.141-151
• /
• 2023

Recently, due to the expansion of the logistics industry, demand for logistics automation equipment is increasing. The modern logistics industry is a high-tech industry that combines various technologies. In general, as various technologies are grafted, the complexity of the system increases, and the occurrence rate of defects and failures also increases. As such, it is time for a predictive maintenance model specialized for logistics automation equipment. In this paper, in order to secure the operational safety and reliability of the parcel loading system, a predictive maintenance platform was implemented based on the Naive Bayes-LSTM(Long Short Term Memory) model. The predictive maintenance platform presented in this paper works by collecting data and receiving data based on a RabbitMQ, loading data in an InMemory method using a Redis, and managing snapshot DB in real time. Also, in this paper, as a verification of the Naive Bayes-LSTM predictive maintenance platform, the function of measuring the time for data collection/storage/processing and determining outliers/normal values was confirmed. The predictive maintenance platform can contribute to securing reliability and safety by identifying potential failures and defects that may occur in the operation of the parcel loading system in the future.