• Nuclear Engineering and Technology
• /
• v.55 no.2
• /
• pp.493-505
• /
• 2023

Comprehensive condition monitoring of large industry systems such as nuclear power plants (NPPs) is essential for safety and maintenance. In this study, we developed novel system-scale diagnostic technology based on deep-learning and IR thermography that can efficiently and cost-effectively classify system conditions using compact Raspberry Pi and IR sensors. This diagnostic technology can identify the presence of an abnormality or accident in whole system, and when an accident occurs, the type of accident and the location of the abnormality can be identified in real-time. For technology development, the experiment for the thermal image measurement and performance validation of major components at each accident condition of NPPs was conducted using a thermal-hydraulic integral effect test facility with compact infrared sensor modules. These thermal images were used for training of deep-learning model, convolutional neural networks (CNN), which is effective for image processing. As a result, a proposed novel diagnostic was developed that can perform diagnosis of components, whole system and accident classification using thermal images. The optimal model was derived based on the modern CNN model and performed prompt and accurate condition monitoring of component and whole system diagnosis, and accident classification. This diagnostic technology is expected to be applied to comprehensive condition monitoring of nuclear power plants for safety.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.603-604
• /
• 2021

It is a enforcement system to prevent collisions caused by unmanageable parking that may occur in parking lots. There are handicapped people who can get up in parking lots, general vehicles parked in electric vehicle parking areas, and vehicles parked in two lanes. The vehicle above is detected and notified through the deep learning object recognition function. By using a picture or video of an unmanageable parking situation as learning data, the learning data is produced so that the situation can be recognized, and the situation is recognized to determine the presence or absence of unmanageable parking. The purpose is to reduce collisions between parking lot users by making the environment of the parking lot more comfortable.

• Smart Structures and Systems
• /
• v.32 no.1
• /
• pp.37-47
• /
• 2023

Indirect determination of pan evaporation (PE) has been highly regarded, due to the advantages of intelligent models employed for this objective. This work pursues improving the reliability of a popular intelligent model, namely multi-layer perceptron (MLP) through surmounting its computational knots. Available climatic data of Fresno weather station (California, USA) is used for this study. In the first step, testing several most common trainers of the MLP revealed the superiority of the Levenberg-Marquardt (LM) algorithm. It, therefore, is considered as the classical training approach. Next, the optimum configurations of two metaheuristic algorithms, namely cuttlefish optimization algorithm (CFOA) and teaching-learning-based optimization (TLBO) are incorporated to optimally train the MLP. In these two models, the LM is replaced with metaheuristic strategies. Overall, the results demonstrated the high competency of the MLP (correlations above 0.997) in the presence of all three strategies. It was also observed that the TLBO enhances the learning and prediction accuracy of the classical MLP (by nearly 7.7% and 9.2%, respectively), while the CFOA performed weaker than LM. Moreover, a comparison between the efficiency of the used metaheuristic optimizers showed that the TLBO is a more time-effective technique for predicting the PE. Hence, it can serve as a promising approach for indirect PE analysis.

• Geomechanics and Engineering
• /
• v.37 no.6
• /
• pp.629-641
• /
• 2024

The presence of excavations or cavities beneath the foundations of a building can have a significant impact on their stability and cause extensive damage. Traditional methods for calculating the bearing capacity and subsidence of foundations over cavities can be complex and time-consuming, particularly when dealing with conditions that vary. In such situations, machine learning (ML) and deep learning (DL) techniques provide effective alternatives. This study concentrates on constructing a prediction model based on the performance of ML and DL algorithms that can be applied in real-world settings. The efficacy of eight algorithms, including Regression Analysis, k-Nearest Neighbor, Decision Tree, Random Forest, Multivariate Regression Spline, Artificial Neural Network, and Deep Neural Network, was evaluated. Using a Python-assisted automation technique integrated with the PLAXIS 2D platform, a dataset containing 272 cases with eight input parameters and one target variable was generated. In general, the DL model performed better than the ML models, and all models, except the regression models, attained outstanding results with an R² greater than 0.90. These models can also be used as surrogate models in reliability analysis to evaluate failure risks and probabilities.

• Journal of The Korean Association of Information Education
• /
• v.26 no.3
• /
• pp.175-186
• /
• 2022

This study has explored the facilitating factors affecting the learning persistence of university students and divided Emergency Remote Teaching(ERT) into two types based on its policies. The conclusions are as follows: First, perceived usefulness was the facilitating factor affecting learning persistence in Synchronous ERT. Therefore, learning persistence would improve by establishing various strategies such as exploring appropriate teaching strategies and building a stable infrastructure. Second, it was perceived usefulness, social presence, and system quality that were the facilitating factors affecting learning persistence in Asynchronous ERT. Thus, learning persistence would increase accompanied by immediate feedback, more active interaction, and so on.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.131-131
• /
• 2021

Deep learning models, especially those based on long short-term memory (LSTM), have presented their superiority in addressing time series data issues recently. This study aims to comprehensively evaluate the performance of deep learning models that belong to the supervised learning category in streamflow prediction. Therefore, six deep learning models-standard LSTM, standard gated recurrent unit (GRU), stacked LSTM, bidirectional LSTM (BiLSTM), feed-forward neural network (FFNN), and convolutional neural network (CNN) models-were of interest in this study. The Red River system, one of the largest river basins in Vietnam, was adopted as a case study. In addition, deep learning models were designed to forecast flowrate for one- and two-day ahead at Son Tay hydrological station on the Red River using a series of observed flowrate data at seven hydrological stations on three major river branches of the Red River system-Thao River, Da River, and Lo River-as the input data for training, validation, and testing. The comparison results have indicated that the four LSTM-based models exhibit significantly better performance and maintain stability than the FFNN and CNN models. Moreover, LSTM-based models may reach impressive predictions even in the presence of upstream reservoirs and dams. In the case of the stacked LSTM and BiLSTM models, the complexity of these models is not accompanied by performance improvement because their respective performance is not higher than the two standard models (LSTM and GRU). As a result, we realized that in the context of hydrological forecasting problems, simple architectural models such as LSTM and GRU (with one hidden layer) are sufficient to produce highly reliable forecasts while minimizing computation time because of the sequential data nature.

• Journal of dental hygiene science
• /
• v.23 no.2
• /
• pp.93-102
• /
• 2023

Background: To investigate political participation by dental hygiene students and analyze the differences therein based on the learning efficacy of dental hygiene policy. Methods: A total of 239 dental hygiene students who were expected to graduate responded to the survey. The data were collected online using a structured questionnaire consisting of 6 items on general characteristics, 10 on political participation, and 15 on the learning efficacy of dental hygiene policy. Statistical analysis was performed using SPSS 23.0. Political participation based on the learning efficacy of dental hygiene policy was analyzed using independent t-tests, ANOVA, and multiple regression analysis (p<0.05). Results: Among the dental hygiene students, 60.7% voted in all three recent presidential, general, and local elections, and 14.2% did not. For political parties supported, 65.7% responded that they had "no supporting party," and 34.3% indicated that they had a "supporting party." In terms of the level of political participation of dental hygiene students (0~50 points), the average score was 25.8 points, with the average passive political participation (0~25 points) score at 15.6 points and the average active political participation (0~25 points) score at 10.2 points. With an increase in dental hygiene policy learning efficacy, both passive and active political participation showed higher scores (p<0.05). Conclusion: Dental hygiene students showed low political participation. The presence of a supporting party, higher voting participation, and higher learning efficacy of dental hygiene policy were associated with higher passive and active political participation. Therefore, to increase this population's interest in political participation, various opportunities for related learning need to be promoted and provided in academia, leading to the enhancement of their political capabilities. In this manner, dental hygienists should expand their capabilities in various roles such as advocates, policy makers, and leaders.

• Informatization Policy
• /
• v.31 no.1
• /
• pp.32-52
• /
• 2024

This study investigates the effectiveness of virtual reality agents as foreign language instructors with focus on the impact of different native language backgrounds and instructional roles. The agents were first distinguished as native or non-native speakers treated as a between-subject factor, and then assigned roles as either teachers or salespersons considered within-subject factors. An immersive virtual environment was developed for this experiment, and a 2×2 mixed factorial design was carried out. In an experimental group of 72 university students, statistically significant interactions were found in learning satisfaction, memory, and recall between the native/non-native status of the agents and their roles. With regard to learning confidence and presence, however, no statistically significant differences were observed in both interaction effects and main effects. Contextual learning in a virtual environment was found to enhance learning effectiveness and satisfaction, with the nativeness and the role of agents influencing learners' memory; thus highlighting the effectiveness of using virtual reality agents in foreign language learning. This suggests that varied approaches can have positive cognitive and emotional impacts on learners, thereby providing valuable theoretical and empirical implications.

• International conference on construction engineering and project management
• /
• 2020.12a
• /
• pp.137-145
• /
• 2020

Providing safety training to construction workers is essential to reduce safety accidents at the construction site. With the prosperity of visualization technologies, Immersive Virtual Reality (IVR) has been adopted for construction safety training by providing interactive learning experiences in a virtual environment. Previous research efforts on IVR-based training have found that the level of fidelity of interaction between real and virtual worlds is one of the important factors contributing to the sense of presence that would affect training performance. Various interactive devices that link activities between real and virtual worlds have been applied in IVR-based training, ranging from existing computer input devices (e.g., keyboard, mouse, joystick, etc.) to specially designed devices such as high-end VR simulators. However, the need for high-fidelity interactive devices may hinder the applicability of IVR-based training as they would be more expensive than IVR headsets. In this regard, this study aims to understand the impact of the level of fidelity of interactive devices in the sense of presence in a virtual environment and the training performance during IVR-based forklift safety training. We conducted a comparative study by recruiting sixty participants, splitting them into two groups, and then providing different interactive devices such as a keyboard for a low fidelity group and a steering wheel and pedals for a high-fidelity group. The results showed that there was no significant difference between the two groups in terms of the sense of presence and task performance. These results indicate that the use of low-fidelity interactive devices would be acceptable for IVR-based safety training as safety training focuses on delivering safety knowledge, and thus would be different from skill transferring training that may need more realistic interaction between real and virtual worlds.

• The Journal of the Korea Contents Association
• /
• v.21 no.1
• /
• pp.292-310
• /
• 2021

The influence of COVID 19 has a direct impact on the education field. In the situation where non-face-to-face classes are inevitably required, interest in the learning satisfaction and intention to use real-time video conferencing is increasing. This study analyzed the effects of learner-teacher interaction, social presence, availability, self-efficacy and academic engagement as video conferencing characteristics and learner's characteristics on perceived usefulness, perceived ease of use, learning satisfaction and intention to use video conferencing. The results of this study showed that learner-teacher interaction, availability, and self-efficacy had a positive effect on perceived ease of use, and all variables except availability had a positive effect on perceived usefulness. Also, perceived usefulness and ease of use were factors that increased learning satisfaction and video conferencing use intention, and learning satisfaction was identified as variables that increased video conferencing use intention. This study has significance in that it provided various theoretical and practical implications for real-time video conferencing which will be used in many educational fields in the future through empirical analysis.