• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.5
• /
• pp.574-583
• /
• 2021

Predicting shipping markets is an important issue. Such predictions form the basis for decisions on investment methods, fleet formation methods, freight rates, etc., which greatly affect the profits and survival of a company. To this end, in this study, we propose a shipping freight rate prediction model for container ships using gated recurrent units (GRUs) and long short-term memory structure. The target of our freight rate prediction is the China Container Freight Index (CCFI), and CCFI data from March 2003 to May 2020 were used for training. The CCFI after June 2020 was first predicted according to each model and then compared and analyzed with the actual CCFI. For the experimental model, a total of six models were designed according to the hyperparameter settings. Additionally, the ARIMA model was included in the experiment for performance comparison with the traditional analysis method. The optimal model was selected based on two evaluation methods. The first evaluation method selects the model with the smallest average value of the root mean square error (RMSE) obtained by repeating each model 10 times. The second method selects the model with the lowest RMSE in all experiments. The experimental results revealed not only the improved accuracy of the deep learning model compared to the traditional time series prediction model, ARIMA, but also the contribution in enhancing the risk management ability of freight fluctuations through deep learning models. On the contrary, in the event of sudden changes in freight owing to the effects of external factors such as the Covid-19 pandemic, the accuracy of the forecasting model reduced. The GRU1 model recorded the lowest RMSE (69.55, 49.35) in both evaluation methods, and it was selected as the optimal model.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.6-8
• /
• 1998

전기실험 관련 인공과학 실험실은 중.고등학교 교과과정의 전기실험을 중심으로 한 임의의 모의 실험을 컴퓨터 상에서 가상적으로 진행 할 수 있도록 한 기존의 인공화학실험실에 기반한 시스템이다. 본 논문에서는 실험 진행을 위해 사용되는 도구 지식을 구조적으로 표현하였으며 실험 진행을 위한 도구간의 공간관계를 정의하였다. 그리고 실험의 전체상태를 나타내는 실험실 상황판의 도구간 관계정보를 통해 생성되는 계산 모델을 설계하였다. 계산 모델은 추론 진행의 조건이 되는 도구의 속성값을 결정하며, 이를 통해 추론을 효율적으로 진행 할 수 있다.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.23 no.6
• /
• pp.466-473
• /
• 2011

The buoyancy and initial momentum fluxes make near-field dominated by buoyant jet when thermal discharge releases underwater. In order to estimate prediction capabilities of those near-field phenomena, non-hydrostatic RANS applied CFD(Computational Fluid Dynamic) model was used. Condition of model was composed based on past laboratory experiments. Numerical simulations carried out for the horizontal buoyant jet in the stagnant flow and vertical buoyant jet into crossflow. The results of simulation are compared with the terms of trajectory and dilution rate of laboratory experiments and analytic model(CorJET) results. CFD model showed a good agreement with them. CFD model can be appropriate for assessment of submerged thermal discharge effect because CFD model can resolve the limitations of near-field analytic model and far-field quasi 3D hydrodynamic model. The accuracy and capability of the CFD model is reviewed in this study. If the computational efficiency get improved, CFD model can be widely applied for simulation of transport and diffusion of submerged thermal discharge.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.36 no.9
• /
• pp.1003-1008
• /
• 2012

The paper presents an Experimental Vehicle Model (EVM), that utilizes the kinematic characteristics of suspensions from SPMD test data. The relative displacement and orientation of a wheel with respect to the body are represented as a function of the vertical displacement of the wheel. The equations of motion of the vehicle are formulated in terms of local coordinates that do not require coordinate transformation, which improves the efficiency of dynamic analysis. The EOM was modularized for each suspension model, and a $6{\times}6$ vehicle model was obtained by combining six suspensions. The analysis results were compared with ADAMS to verify the accuracy of the EVM. This study also verifies the feasibility of real-time simulation with the developed EVM. For a vehicle simulation for 1 ms, the real simulation time required within 20% of the prescribed time. This result shows that the EVM meets the real-time simulation requirements.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.2A
• /
• pp.285-292
• /
• 2006

In this study, a method determining the local bond-slip model from pure shear test results of CFRP sheet-concrete adhesive joints is proposed and local bond-slip models are presented. Adhesive joints with a specific bond-slip model, which is assumed as multi-linear curve in order to represent arbitary function, are solved numerically. The difference between the solution and test results are minimized for finding the bond-slip model. The model with bilinear curve is also optimized to verify the improvement of multi-linear model. The selected test results are ultimate load-adhesive length curves from a series of adhesive joints and load-displacement curves for each joint. The optimization problem is formulated by physical programming, and the optimized bond-slip model is found using genetic algorithm.

• Korean Journal of Cognitive Science
• /
• v.10 no.1
• /
• pp.57-66
• /
• 1999

We believe that science experiments in a laboratory are essential for science education. Scientific experiments begin with situations set by selecting and locating tools and reagents. and by proper experimental behavior, and thereafter situations are changed by natural laws and intermediate experimental behavior. While scientists and students do experiments, they build a cognitive model internally, do causal reasoning on the model to derive system behavior, and then learn scientific truth. We suggest not only a representation method for a 2-dimentional model and for ontological entities necessary in causal reasoning, but also an inferencing method to derive behavior. Chemistry experiments are chosen for the implementation. For the ontological entities, we consider experimental tools, reagents and their heirarchical structures, physics and chemistry natural laws, and functional abstraction knowledge. In order to show the usefulness of our methods, we have developed a program, called ACUArtificial Chemistry Laboratory), which provides an experiment environment where students can do non-predetermined experiments, and shows experiment려 system behavior similar to what happens in the same situation in a real world and descriptions about why it happens.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.23 no.2
• /
• pp.123-137
• /
• 2010

The hybrid structural experiment decomposes a structure into independent substructures that can be tested or simulated. The substructures being tested or simulated may be distributed at different facilities of different locations, and are managed by the simulation coordinator. There exist interactions among the simulation coordinator and the substructures since they give and receive the commands and feedbacks during the experimental process. These interactions are described in this paper for an example hybrid structural experiment using the classes and objects in the Lehigh Model which is one of the data models for structural experiments. The simulation coordinator and the substructures have the objects for the interaction data files, and are linked together through the same types of the interface links. The objects for the interactions presented in this paper can be implemented in a consistent way, and be used for developing the computer system for the hybrid structural experiments.

• The Journal of the Korea Contents Association
• /
• v.13 no.4
• /
• pp.53-63
• /
• 2013

This paper presents an importance sampling(IS) embedded experimental frame(EF) design for efficient Monte Carlo (MC) simulation. To achieve IS principles, the proposed EF contains two embedded sub-models, which are classified into Importance Sampler(IS) and Bias Compensator(BC) models. The IS and BC models stand between the existing system model and EF, which leads to enhancement of model reusability. Furthermore, the proposed EF enables to achieve fast stochastic simulation as compared with the crude MC technique. From the abstract two case studies with the utilization of the proposed EF, we can gain interesting experimental results regarding remarkable enhancement of simulation performance. Finally, we expect that this work will serve various content areas for enhancing simulation performance, and besides, it will be utilized as a tool to understand and analyze social phenomena.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.17 no.3
• /
• pp.98-110
• /
• 2018

In this paper, we propose both Dense grid model and Anchor model to improve the recognition rate of dynamic objects. Two experiments are conducted to study the performance of two proposed CNNs models (Dense grid model and Anchor model), which are to detect dynamic objects. In the first experiment, YOLO-v2 network is adjusted, and then fine-tuned on KITTI datasets. The Dense grid model and Anchor model are then compared with YOLO-v2. Regarding to the evaluation, the two models outperform YOLO-v2 from 6.26% to 10.99% on car detection at different difficulty levels. In the second experiment, this paper conducted further training of the models on a new dataset. The two models outperform YOLO-v2 up to 22.40% on car detection at different difficulty levels.

• Journal of the Korean Society for information Management
• /
• v.23 no.1 s.59
• /
• pp.201-219
• /
• 2006

This experimental study suggests an element-based XML document retrieval method that reveals highly relevant elements. The models investigated here for comparison are divergence and smoothing method, and hierarchical language model. In conclusion, the hierarchical language model proved to be most effective in element-based XML document retrieval with regard to the improved exhaustivity and harmed specificity.