• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.1
• /
• pp.13-21
• /
• 2021

The purpose of this study is to develop the ship valuation model by utilizing the neural network model. The target of the valuation was secondhand VLCC. The variables were set as major factors inducing changes in the value of ship through prior research, and the corresponding data were collected on a monthly basis from January 2000 to August 2020. To determine the stability of subsequent variables, a multi-collinearity test was carried out and finally the research structure was designed by selecting six independent variables and one dependent variable. Based on this structure, a total of nine simulation models were designed using linear regression, neural network regression, and random forest algorithm. In addition, the accuracy of the evaluation results are improved through comparative verification between each model. As a result of the evaluation, it was found that the most accurate when the neural network regression model, which consist of a hidden layer composed of two layers, was simulated through comparison with actual VLCC values. The possible implications of this study first, creative research in terms of applying neural network model to ship valuation; this deviates from the existing formalized evaluation techniques. Second, the objectivity of research results was enhanced from a dynamic perspective by analyzing and predicting the factors of changes in the shipping. market.

• Nuclear Engineering and Technology
• /
• v.54 no.1
• /
• pp.162-176
• /
• 2022

In Lead-based reactor (LBR) severe accident, the meltdown and migration inside the reactor core will lead to fuel fragment concentration, which may further cause re-criticality and even core disintegration. Accurately predicting the migration and solidification behavior of melt in LBR severe accidents is of prime importance for safety analysis of LBR. In this study, the Moving Particle Semi-implicit (MPS) method is validated and used to simulate the migration and solidification behavior. Two main surface tension models are validated and compared. Meanwhile, the MPS method is validated by the L-plate solidification test. Based on the improved MPS method, the migration and solidification behavior of melt in LBR severe accident was studied furthermore. In the Pb-Bi coolant, the melt flows upward due to density difference. The migration and solidification behavior are greatly affected by the surface tension and viscous resistance varying with enthalpy. The whole movement process can be divided into three stages depending on the change in velocity. The heat transfer of core melt is determined jointly by two heat transfer modes: flow heat transfer and solid conductivity. Generally, the research results indicate that the MPS method has unique advantage in studying the migration and solidification behavior in LBR severe accident.

• Journal of Practical Agriculture & Fisheries Research
• /
• v.24 no.2
• /
• pp.12-22
• /
• 2022

The purpose of this study is to design a farming model for Hanwoo start-up farmers. I prepared a Hanwoo production plan model according to the growth cycle of Hanwoo using EXCEL. The Hanwoo production plan model was simulated in two model: Model 1 (a model that only purchases Hanwoo calf) and Model 2 (a model that purchases both Hanwoo cow and Hanwoo calf). Next, I reviewed the profits and costs of two Hanwoo simulation models. As a result of the analysis, Model 2 has the following characteristics compared to Model 1. First, Model 2 requires a lot of initial investment. Second, Model 2 is advantageous in terms of farm cash balance because imports occur every year. Third, Model 2 can efficiently use facilities and machines.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.50 no.8
• /
• pp.531-540
• /
• 2022

In this paper, we present an algorithm that identifies artillery type and rapidly predicts the impact point based on the IMM filter. The ballistic trajectory equation is used as a system model, and three models with different ballistic coefficient values are used. Acceleration was divided into three components of gravity, air resistance, and lift. And lift acceleration was added as a new state variable. The kinematic condition that the velocity vector and lift acceleration are perpendicular was used as a pseudo-measurement value. The impact point was predicted based on the state variable estimated through the IMM filter and the ballistic coefficient of the model with the highest mode probability. Instead of the commonly used Runge-Kutta numerical integration for impact point prediction, a semi-analytic method was used to predict impact point with a small amount of calculation. Finally, a state variable initialization method using the least-square method was proposed. An integrated algorithm including artillery type identification, impact point prediction and initialization was presented, and the validity of the proposed method was verified through simulation.

• Journal of Korea Water Resources Association
• /
• v.54 no.7
• /
• pp.485-493
• /
• 2021

Groundwater is an important water resource that can be used along with surface water. In particular, in the case of island regions, research on groundwater level variability is essential for stable groundwater use because the ratio of groundwater use is relatively high. Researches using artificial intelligence models (AIs) for the prediction and analysis of groundwater level variability are continuously increasing. However, there are insufficient studies presenting evaluation criteria to judge the appropriateness of groundwater level prediction. This study comprehensively analyzed the research results that predicted the groundwater level using AIs for various regions around the world over the past 20 years to present the range of allowable groundwater level prediction errors. As a result, the groundwater level prediction error increased as the observed groundwater level variability increased. Therefore, the criteria for evaluating the adequacy of the groundwater level prediction by an AI is presented as follows: less than or equal to the root mean square error or maximum error calculated using the linear regression equations presented in this study, or NSE ≥ 0.849 or R² ≥ 0.880. This allowable prediction error range can be used as a reference for determining the appropriateness of the groundwater level prediction using an AI.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.25 no.6
• /
• pp.51-64
• /
• 2022

Along with the urbanization, the risk of urban flooding due to climate change is increasing. Flood regulation, one of the ecosystem services, is implemented in the different level of function of flood risk mitigation by the type of ecosystem such as forests, arable land, wetlands etc. Land use changes due to development pressures have become an important factor in increasing the vulnerability by flash flood. This study has conducted evaluating the urban flood regulation service using InVEST UFRM(Urban Flood Risk Model). As a result of the simulation, the potential water retention by ecosystem type in the event of a flash flood according to RCP 4.5(10 year frequency) scenario was 1,569,611 tons in urbanized/dried areas, 907,706 tons in agricultural areas, 1,496,105 tons in forested areas, 831,705 tons in grasslands, 1,021,742 tons in wetlands, and 206,709 tons in bare areas, the water bodies was estimated to be 38,087 tons. In the case of more severe 100-year rainfall, 1,808,376 tons in urbanized/dried areas, 1,172,505 tons in agricultural areas, 2,076,019 tons in forests, 1,021,742 tons in grasslands, 47,603 tons in wetlands, 238,363 tons in bare lands, and 52,985 tons in water bodies. The potential economic damage from flood runoff(100 years frequency) is 122,512,524 thousand won in residential areas, 512,382,410 thousand won in commercial areas, 50,414,646 thousand won in industrial areas, 2,927,508 thousand won in Infrastructure(road), 8,907 thousand won in agriculture, Total of assuming a runoff of 50 mm(100 year frequency) was estimated at 688,245,997 thousand won. In a conclusion. these results provided an overview of ecosystem functions and services in terms of flood control, and indirectly demonstrated the possibility of using the model as a tool for policy decision-making. Nevertheless, in future research, related issues such as application of models according to various spatial scales, verification of difference in result values due to differences in spatial resolution, improvement of CN(Curved Number) suitable for the research site conditions based on actual data, and development of flood damage factors suitable for domestic condition for the calculation of economic loss.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.11
• /
• pp.449-456
• /
• 2021

An intrusion detection system is a technology that detects abnormal behaviors that violate security, and detects abnormal operations and prevents system attacks. Existing intrusion detection systems have been designed using statistical analysis or anomaly detection techniques for traffic patterns, but modern systems generate a variety of traffic different from existing systems due to rapidly growing technologies, so the existing methods have limitations. In order to overcome this limitation, study on intrusion detection methods applying various machine learning techniques is being actively conducted. In this study, a comparative study was conducted on data preprocessing techniques that can improve the accuracy of anomaly detection using NGIDS-DS (Next Generation IDS Database) generated by simulation equipment for traffic in various network environments. Padding and sliding window were used as data preprocessing, and an oversampling technique with Adversarial Auto-Encoder (AAE) was applied to solve the problem of imbalance between the normal data rate and the abnormal data rate. In addition, the performance improvement of detection accuracy was confirmed by using Skip-gram among the Word2Vec techniques that can extract feature vectors of preprocessed sequence data. PCA-SVM and GRU were used as models for comparative experiments, and the experimental results showed better performance when sliding window, skip-gram, AAE, and GRU were applied.

• Journal of Korean Elementary Science Education
• /
• v.41 no.3
• /
• pp.538-552
• /
• 2022

This study attempted to understand the characteristics of pedagogic activities performed by pre-service elementary school teachers. To this end, it applied Algodoo simulations to analyze the actions of students and obtain educational implications for optical learning. The study's participants comprised 79 first-year students enrolled in a teacher training college. Their activities could be classified as representation reproductions, verification experiments, and inquiry experiments. Students who performed representation reproduction exercises replicated renowned and authoritative exemplars, apprehending and demonstrating their principal features through simulations. Students performing verification experiments attempted to validate previously learned optical concepts by reviewing the relevant theoretical contexts. Such students primarily conducted simple experiments. Students accomplishing inquiry experiments used simulations to explore phenomena they did not know. Some of them even investigated optical phenomena beyond the domain of general physics. The above results confirmed that free optical experiments performed using Algodoo can effectively denote starting points for learners to engage in activities at varying levels. Additionally, students require assistance from instructors in addressing queries about the application of the principles and models related to optics. This study suggests ways in which instructors should help students at each level of activity. Additionally, the paper presents examples of varying levels of inquiry-related activities available on Algodoo. It also discusses the advantages and disadvantages of performing inquiry-based activities on Algodoo and suggests ways of enhancing the learning achieved through this platform.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.18 no.1
• /
• pp.45-52
• /
• 2023

The power generation method using expensive diesel has operation problems such as high cost diesel generator and a lack of reserved power due to increase of power demand in some islands, requiring expansion of power generation facilities. To solve this problems, it is necessary to improve the efficiency of power generation facilities through an ORC(Organic Rankin Cycle) system application that uses waste heat as a heat source. Therefore, localized application technology of price competitive and highly reliable ORC power generation system is needed, and optimization technology of generators is having great effect, so this study performed two generator designs to get a high-efficiency generator with an optimized 30kW output. The comparison of simulation data for two designed models showed that a generator with SPM factor of 46.2% had an efficiency of 92.1% and a power ouput of about 23.2kW based on 12,000rpm, a generator with SPM factor of 44.46%, had a power output of 27.9kW and efficiency of 93.6% based on above rpm. For the verification of improved design model with SPM factor of 44.46%, the prototype test system with 110kW motor dynamometer was installed and got to the efficiency of 92.08% with conditions of the rated capacity 25kW at 12,000rpm, the test results of prototype generator showed the validity of generator design.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.9
• /
• pp.407-418
• /
• 2023

The Multi-Object Goal Visual Navigation(MultiOn) is a visual navigation task in which an agent must visit to multiple object goals in an unknown indoor environment in a given order. Existing models for the MultiOn task suffer from the limitation that they cannot utilize an integrated view of multimodal context because use only a unimodal context map. To overcome this limitation, in this paper, we propose a novel deep neural network-based agent model for MultiOn task. The proposed model, MCFMO, uses a multimodal context map, containing visual appearance features, semantic features of environmental objects, and goal object features. Moreover, the proposed model effectively fuses these three heterogeneous features into a global multimodal context map by using a point-wise convolutional neural network module. Lastly, the proposed model adopts an auxiliary task learning module to predict the observation status, goal direction and the goal distance, which can guide to learn the navigational policy efficiently. Conducting various quantitative and qualitative experiments using the Habitat-Matterport3D simulation environment and scene dataset, we demonstrate the superiority of the proposed model.