• Transactions of Materials Processing
• /
• v.26 no.3
• /
• pp.137-143
• /
• 2017

In this study, a wear test method was proposed to predict the wear life of the CrN layer coated on the surface of the press tools for manufacturing the auto-parts with ultra high strength steel (UHSS) with a tensile strength of 1.5 GPa. The pin-on-disc type wear test was carried out to confirm the feasibility and the reproducibility of the wear amount according to the test conditions such as the normal force, the sliding velocity, and the sliding speed. The test conditions were obtained from the finite element stamping analysis and the wear simulation. With the wear amount from the wear test, a prediction model of the wear depth in the CrN coating layer was proposed according to the test conditions with the design of experiments such as Taguchi method and the response surface method. The derived prediction model was then compared to the result of the Archard wear model, fully describing that the proposed model can effectively predict the wear life of the press tools for the auto-parts with UHSS.

• Journal of Multimedia Information System
• /
• v.8 no.3
• /
• pp.147-158
• /
• 2021

Versatile Video Coding (VVC) is the latest video coding standard developed by Joint Video Exploration Team (JVET). In VVC, the quadtree plus multi-type tree (QT+MTT) structure of coding unit (CU) partition is adopted, and its computational complexity is considerably high due to the brute-force search for recursive rate-distortion (RD) optimization. In this paper, we aim to reduce the time complexity of inter-picture prediction mode since the inter prediction accounts for a large portion of the total encoding time. The problem can be defined as classifying the split mode of each CU. To classify the split mode effectively, a novel convolutional neural network (CNN) called multi-level tree (MLT-CNN) architecture is introduced. For boosting classification performance, we utilize additional information including inter-picture information while training the CNN. The overall algorithm including the MLT-CNN inference process is implemented on VVC Test Model (VTM) 11.0. The CUs of size 128×128 can be the inputs of the CNN. The sequences are encoded at the random access (RA) configuration with five QP values {22, 27, 32, 37, 42}. The experimental results show that the proposed algorithm can reduce the computational complexity by 11.53% on average, and 26.14% for the maximum with an average 1.01% of the increase in Bjøntegaard delta bit rate (BDBR). Especially, the proposed method shows higher performance on the sequences of the A and B classes, reducing 9.81%~26.14% of encoding time with 0.95%~3.28% of the BDBR increase.

• Nuclear Engineering and Technology
• /
• v.53 no.12
• /
• pp.3892-3901
• /
• 2021

The pressure drop of a moisture separator in a steam generator is the important design parameter to ensure the successful performance of a nuclear power plant. The moisture separators have a wide range of operating conditions based on the arrangement of them. The prediction of the pressure drop in a moisture separator is challenging due to the complexity of the multi-dimensional two-phase vortex flow. In this study, the moisture separator test facility using the air/water two-phase flow was used to predict the pressure drop of a moisture separator in a Korean OPR-1000 reactor. The prototypical steam/water two-phase flow conditions in a steam generator were simulated as air/water two-phase flow conditions by preserving the centrifugal force and vapor quality. A series of experiments were carried out to investigate the effect of hydraulic characteristics such as the quality and liquid mass flux on the two-phase pressure drop. A new prediction model based on the scaling law was suggested and validated experimentally using the full and half scale of separators. The suggested prediction model showed good agreement with the steam/water experimental results, and it can be extended to predict the steam/water two-phase pressure drop for moisture separators.

• The Journal of Information Systems
• /
• v.30 no.2
• /
• pp.105-126
• /
• 2021

Purpose The purpose of this study is to implement a optimal machine learning model about the cancellation prediction performance in car sales business. It is to apply the data set of accumulated contract, cancellation, and sales information in sales support system(SFA) which is commonly used for sales, customers and inventory management by imported car dealers, to several machine learning models and predict performance of cancellation. Design/methodology/approach This study extracts 29,073 contracts, cancellations, and sales data from 2015 to 2020 accumulated in the sales support system(SFA) for imported car dealers and uses the analysis program Python Jupiter notebook in order to perform data pre-processing, verification, and modeling that is applying and learning to Machine learning model after then the final result was predicted using new data. Findings This study confirmed that cancellation prediction is possible by applying car purchase contract information to machine learning models. It proved the possibility of developing and utilizing a generalized predictive model by using data of imported car sales system with machine learning technology. It can reduce and prevent the sales failure as caring the potential lost customer intensively and it lead to increase sales revenue by predicting the cancellation possibility of individual customers.

• Food Science of Animal Resources
• /
• v.42 no.4
• /
• pp.655-671
• /
• 2022

There are differences of spectral characteristics between different types of meat cut, which means the model established using only one type of meat cut for meat quality prediction is not suitable for other meat cut types. A novel portable visible and near-infrared (Vis/NIR) optical system was used to simultaneously predict multiple quality indicators for different commercial meat cut types (silverside, back strap, oyster, fillet, thick flank, and tenderloin) from Small-tailed Han sheep. The correlation coefficients of the calibration set (R_c) and prediction set (R_p) of the optimal prediction models were 0.82 and 0.81 for pH, 0.88 and 0.84 for L^*, 0.83 and 0.78 for a^*, 0.83 and 0.82 for b^*, 0.94 and 0.86 for cooking loss, 0.90 and 0.88 for shear force, 0.84 and 0.83 for protein, 0.93 and 0.83 for fat, 0.92 and 0.87 for moisture contents, respectively. This study demonstrates that Vis/NIR spectroscopy is a promising tool to achieve the predictions of multiple quality parameters for different commercial meat cut types.

• International journal of advanced smart convergence
• /
• v.12 no.1
• /
• pp.53-58
• /
• 2023

With the fast development of artificial intelligence day by day, users are demanding explanations about the results of algorithms and want to know what parameters influence the results. In this paper, we propose a model for bankruptcy prediction with interpretability using the SHAP framework. SHAP (SHAPley Additive exPlanations) is framework that gives a visualized result that can be used for explanation and interpretation of machine learning models. As a result, we can describe which features are important for the result of our deep learning model. SHAP framework Force plot result gives us top features which are mainly reflecting overall model score. Even though Fully Connected Neural Networks are a "black box" model, Shapley values help us to alleviate the "black box" problem. FCNNs perform well with complex dataset with more than 60 financial ratios. Combined with SHAP framework, we create an effective model with understandable interpretation. Bankruptcy is a rare event, then we avoid imbalanced dataset problem with the help of SMOTE. SMOTE is one of the oversampling technique that resulting synthetic samples are generated for the minority class. It uses K-nearest neighbors algorithm for line connecting method in order to producing examples. We expect our model results assist financial analysts who are interested in forecasting bankruptcy prediction of companies in detail.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.28 no.1
• /
• pp.175-183
• /
• 2022

The collision between a ship and bridge across a waterway may result in extremely serious consequences that may endanger the safety of life and property. Therefore, factors affecting ship bridge collision must be investigated, and the impact force should be discussed based on various collision conditions. In this study, a finite element model of ship bridge collision is established, and the peak impact force of a ship bridge collision based on 50 operating conditions combined with three input parameters, i.e., ship loading condition, ship speed, and ship bridge collision angle, is calculated via numerical simulation. Using neural network models trained with the numerical simulation results, the prediction model of the peak impact force of ship bridge collision involving an extremely short calculation time on the order of milliseconds is established. The neural network models used in this study are the basic backpropagation neural network model and Elman neural network model, which can manage temporal information. The accuracy of the neural network models is verified using 10 test samples based on the operating conditions. Results of a verification test show that the Elman neural network model performs better than the backpropagation neural network model, with a mean relative error of 4.566% and relative errors of less than 5% in 8 among 10 test cases. The trained neural network can yield a reliable ship bridge collision force instantaneously only when the required parameters are specified and a nonlinear finite element solution process is not required. The proposed model can be used to predict whether a catastrophic collision will occur during ship navigation, and thus hence the safety of crew operating the ship.

• Journal of Drive and Control
• /
• v.20 no.4
• /
• pp.115-122
• /
• 2023

This study utilized a discrete element method (DEM) simulation, as one of the virtual field trials, to predict the impact of tillage depth on the rotary blade shaft during rotavator tilling. The virtual field for the simulation was generated according to soil properties observed in an actual field. Following the generation of particles for the virtual field, a sequence of calibration steps followed to align the mechanical properties more closely with those of real soil. Calibration was conducted with a focus on bulk density and shear torque, resulting in calibration errors of just 0.02% for bulk density and 0.52% for shear torque. The prediction of the load on a rotary tiller's blade shaft involved a three-pronged approach, considering shaft torque, draft force, and vertical force. In terms of shaft torque, the values exhibited significant increases of 42.34% and 36.91% for every 5-centimeter increment in tillage depth. Similarly, the vertical force saw substantial growth by 40.41% and 36.08% for every 5-centimeter increment. In contrast, the variation in draft force based on tillage depth was comparatively lower at 18.49% and 0.96%, indicating that the effect of tillage depth on draft force was less pronounced than its impact on shaft torque and vertical force. From a perspective of agricultural machinery research, this study provides valuable insights into the DEM soil modeling process, accounting for changes in soil properties with varying tillage depths. These findings are expected to be instrumental in future agricultural machinery design studies.

• Journal of Biosystems Engineering
• /
• v.14 no.2
• /
• pp.67-79
• /
• 1989

This study was conducted to investigate the applicability of true model theory for pull prediction in a powered lugged wheel-soil system and to examine the possibility of using principles of similitude in investigating the dimensions of soil parameters pertinent to a powered lugged wheel-soil system concerning the pull prediction. The following conclusions were derived from the study; 1) The pull of prototype wheels proved to be predicted by those of the model wheels for the range of the dynamic weight tested. The pull curves of models and prototype were respectively very similar in the shape. From this basic knowledge, it was enabled to apply the similitude theory to the performance prediction of the true model. 2) A conditional equation which can be used for the prediction of pull of prototype by model test was derived as follows. $n_f=n_1^{-b}$ where $n_f$ : force scale = $w/w_m$ $n_1$ : length scale = ${\ell}/{\ell}_m$ b : exponent on the length dimension of the soil property ${\alpha}$ The range of the numerical value of b, which was determined by the least square method, was found to be -2.0~-2.6. 3) Considering a relatively wide variation of b values in the pull prediction, b is considered to be a function of many variales. Thus it was concluded that there are several soil properties which are pertinent to the powered lugged-wheel-soil system concerning the pull prediction, and these soil properties may have the different effects on the pull of model and protytype wheels, to give the different dimension on the soil parameters.

• Food Engineering Progress
• /
• v.14 no.1
• /
• pp.75-79
• /
• 2010

Partial least square regression (PLSR) was executed on curve data of force-deformation from back extrusion test and sensory data for commercial instant noodles. Sensory attributes considered were hardness (A), springiness (B), roughness (C), adhesiveness to teeth (D), and thickness (E). Eight and two kinds of fried and non-fried instant noodles respectively were used in the tests. Changes in weighted regression coefficients were characterized as three stages: compaction, yielding, and extrusion. Correlation coefficients appeared in the order of E>D>A>B>C, root mean square error of prediction D>C>E>B>A, and relative ability of prediction D>C>E>B>A. Overall, 'D' was the best in the correlation and prediction. 'A' with poor prediction ability but high correlation was considered good when determining the order of magnitude.