• Journal of Korea Water Resources Association
• /
• v.53 no.6
• /
• pp.395-408
• /
• 2020

This paper aims to develop and apply three different machine learning approaches (i.e., artificial neural networks (ANN), adaptive neuro-fuzzy inference systems (ANFIS), and wavelet-based neural networks (WNN)) combined with an evolutionary optimization algorithm and the k-fold cross validation for multi-step (days) streamflow forecasting at the catchment located in Algeria, North Africa. The ANN and ANFIS models yielded similar performances, based on four different statistical indices (i.e., root mean squared error (RMSE), Nash-Sutcliffe efficiency (NSE), correlation coefficient (R), and peak flow criteria (PFC)) for training and testing phases. The values of RMSE and PFC for the WNN model (e.g., RMSE = 8.590 ㎥/sec, PFC = 0.252 for (t+1) day, testing phase) were lower than those of ANN (e.g., RMSE = 19.120 ㎥/sec, PFC = 0.446 for (t+1) day, testing phase) and ANFIS (e.g., RMSE = 18.520 ㎥/sec, PFC = 0.444 for (t+1) day, testing phase) models, while the values of NSE and R for WNN model were higher than those of ANNs and ANFIS models. Therefore, the new approach can be a robust tool for multi-step (days) streamflow forecasting in the Seybous River, Algeria.

• Korean Chemical Engineering Research
• /
• v.46 no.1
• /
• pp.194-199
• /
• 2008

In this study, the effective method to esterify the free fatty acids in jatropha oil was examined. Compared to other plant oils, the acid value of jatropha oil was remarkably high, 11.5 mgKOH/g. So direct transesterification by a base catalyst was not suitable for the oil. After the free fatty acids were esterified with methanol, jatropha oil was transesterified. The activities of four solid acid catalysts were tested and Amberlyst-15 showed the best activity for the esterification. After constructing the experiment matrix based on RSM and analyzing the statistical data, the optimal esterification conditions were determined to be 6.79% of methanol and 17.14% of Amberlyst-15. After the pretreatment, jatropha biodiesel was produced by the transesterification using KOH in a pressurized batch reactor. Jatropha biodiesel produced could meet the major specifications of Korean biodiesel standards; 97.35% of FAME, 8.17 h of oxidation stability, 0.125% of total glycerol and $0^{\circ}C$ of CFPP.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.34 no.3
• /
• pp.307-313
• /
• 2010

In this study, alternative designs for the bends used in district heating pipes are investigated. The district heating pipes, which are subjected to temperatures of 10 to $120^{\circ}C$ and a water pressure of $16\;kgf/cm^2$, have to withstand thermomechanical cyclic loads when in use. These pipes comprise three concentric tubes: a steel pipe (internal), polyurethane (PUR) insulator (middle), and a high-density polyethylene (HDPE) case (external). In addition, the bends in the district heating pipe system are covered with foam pads that cause aging. In this study, an alternative bend design that does not involve the use of a foam pad is proposed to overcome the aging problem in the bends. In the proposed design, "shear rings" are added to the surface of a bend, and its dimensions are determined by a combination of the statistical (Taguchi) method and FEM. The geometrical parameters such as thickness, height, and number of the rings significantly affect the design optimization, and hence, they affect the results of the FEM.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.2
• /
• pp.385-391
• /
• 2000

Recently, neural network research for forecasting the consecutive controlling rules of the future is being progressed, using the series data which are different from the traditional statistical analysis methods. In this paper, we suggest the pruning algorithm for the fast and exact weather forecast that excludes the hidden layer of the early optional designed nenral network. There are perform the weather forecast experiments using the 22080 kinds of weather data gathered from 1987 to 1996 for proving the efficiency of this suggested algorithm. Through the experiments, the early optional composed $26{\times}50{\times}1$ nenral network became the most suitable $26{\times}2{\times}1$ structure through the pruning algorithm suggested, in the optimum neural network $26{\times}2{\times}1$, in the case of the error temperature ${\pm}0.5^{\circ}C$, the average was 33.55%, in the case of ${\pm}1^{\circ}C$, the average was 61.57%, they showed more superior than the average 29.31% and 54.47% of the optional designed structure, also. we can reduce the calculation frequency more than maximum 25 times as compared with the optional sturcture neural network in the calculation frequencies.

• Journal of Korean Neurosurgical Society
• /
• v.61 no.6
• /
• pp.753-760
• /
• 2018

Objective : We investigated the effect of optimization in dose-limiting shell method on the dosimetric quality of CyberKnife (CK) plans in treating brain metastases (BMs). Methods : We selected 19 BMs previously treated using CK between 2014 and 2015. The original CK plans ($CK_{original}$) had been produced using 1 to 3 dose-limiting shells : one at the prescription isodose level (PIDL) for dose conformity and the others at low-isodose levels (10-30% of prescription dose) for dose spillage. In each case, a modified CK plan ($CK_{modified}$) was generated using 5 dose-limiting shells : one at the PIDL, another at intermediate isodose level (50% of prescription dose) for steeper dose fall-off, and the others at low-isodose levels, with an optimized shell-dilation size based on our experience. A Gamma Knife (GK) plan was also produced using the original contour set. Thus, three data sets of dosimetric parameters were generated and compared. Results : There were no differences in the conformity indices among the $CK_{original}$, $CK_{modified}$, and GK plans (mean 1.22, 1.18, and 1.24, respectively; p=0.079) and tumor coverage (mean 99.5%, 99.5%, and 99.4%, respectively; p=0.177), whereas the $CK_{modified}$ plans produced significantly smaller normal tissue volumes receiving 50% of prescription dose than those produced by the $CK_{original}$ plans (p<0.001), with no statistical differences in those volumes compared with GK plans (p=0.345). Conclusion : These results indicate that significantly steeper dose fall-off is able to be achieved in the CK system by optimizing the shell function while maintaining high conformity of dose to tumor.

• Korean Journal of Food Science and Technology
• /
• v.53 no.6
• /
• pp.739-748
• /
• 2021

This study focuses on the use of rice in the production of gluten-free rice udon (GFU) through an optimized mixing ratio, using the Box-Behnken response surface methodology (RSM). Different additional levels of rice flour (A, 40-60 g), acetylated distarch adipate (B, 10-20 g), and trehalose (C, 0-3 g) were used as variables, while water absorption level, volume, cooking loss, solid yield, lightness, texture properties, proximate compositions of GFU and turbidity of cooking water were set as responses in the RSM design model. The optimum mixing ratio for the preparation of gluten-free rice udon was obtained for 60.00 g of rice flour, 18.81 g of acetylated distarch adipate without the addition of trehalose. The response values of the optimized samples were water absorption (60.94%), volume (34.94%), turbidity of the cooking water (0.37), cooking loss (4.77%), solid yield (1.55 g), lightness value (70.04), hardness (2.53 N), springiness (0.18), gumminess (10.45 N), chewiness (1.83 N), and cohesiveness (2.89). This study has shown that rice flour can replace wheat flour to manufacture udon at an optimized mixing ratio successfully derived by statistical estimation method.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.34 no.1
• /
• pp.25-33
• /
• 2021

With the development of machine learning techniques, various types of data such as vibration, temperature, and flow rate can be used to detect and diagnose abnormalities in machine conditions. In particular, in the field of the state monitoring of rotating machines, the fault diagnosis of machines using vibration data has long been carried out, and the methods are also very diverse. In this study, an experiment was conducted to collect vibration data from normal and abnormal compressors by installing accelerometers directly on rotary compressors used in household air conditioners. Data segmentation was performed to solve the data shortage problem, and the main features for the fault classification model were extracted through the chi-square test after statistical and physical features were extracted from the vibration data in the time domain. The support vector machine (SVM) model was developed to classify the normal or abnormal conditions of compressors and improve the classification accuracy through the hyperparameter optimization of the SVM.

• Journal of Navigation and Port Research
• /
• v.46 no.3
• /
• pp.280-288
• /
• 2022

This study examined the diagnostics of abnormalities and faults of equipment, whose rotational speed changes even during regular operation. The purpose of this study was to suggest a procedure that can properly apply machine learning to the time series data, comprising non-stationary characteristics as the rotational speed changes. Anomaly and fault diagnosis was performed using machine learning: k-Nearest Neighbor (k-NN), Support Vector Machine (SVM), and Random Forest. To compare the diagnostic accuracy, an autoencoder was used for anomaly detection and a convolution based Conv1D was additionally used for fault diagnosis. Feature vectors comprising statistical and frequency attributes were extracted, and normalization & dimensional reduction were applied to the extracted feature vectors. Changes in the diagnostic accuracy of machine learning according to feature selection, normalization, and dimensional reduction are explained. The hyperparameter optimization process and the layered structure are also described for each algorithm. Finally, results show that machine learning can accurately diagnose the failure of a variable-rotation machine under the appropriate feature treatment, although the convolution algorithms have been widely applied to the considered problem.

• Journal of Microbiology and Biotechnology
• /
• v.32 no.5
• /
• pp.630-637
• /
• 2022

The objective of this study was to optimize industrial-grade media for improving the biomass production of Weissella cibaria JW15 (JW15) using a statistical approach. Eleven variables comprising three carbon sources (glucose, fructose, and sucrose), three nitrogen sources (protease peptone, yeast extract, and soy peptone), and five mineral sources (K₂HPO₄, potassium citrate, ⳑ-cysteine phosphate, MgSO₄, and MnSO₄) were screened by using the Plackett-Burman design. Consequently, glucose, sucrose, and soy peptone were used as significant variables in response surface methodology (RSM). The composition of the optimal medium (OM) was 22.35 g/l glucose, 15.57 g/l sucrose, and 10.05 g/l soy peptone, 2.0 g/l K₂HPO₄, 5.0 g/l sodium acetate, 0.1 g/l MgSO₄·7H₂O, 0.05 g/l MnSO₄·H₂O, and 1.0 g/l Tween 80. The OM significantly improved the biomass production of JW15 over an established commercial medium (MRS). After fermenting OM, the dry cell weight of JW15 was 4.89 g/l, which was comparable to the predicted value (4.77 g/l), and 1.67 times higher than that of the MRS medium (3.02 g/l). Correspondingly, JW15 showed a rapid and increased production of lactic and acetic acid in the OM. To perform a scale-up validation, batch fermentation was executed in a 5-l bioreactor at 37℃ with or without a pH control at 6.0 ± 0.1. The biomass production of JW15 significantly improved (1.98 times higher) under the pH control, and the cost of OM was reduced by two-thirds compared to that in the MRS medium. In conclusion, OM may be utilized for mass producing JW15 for industrial use.

• Wind and Structures
• /
• v.34 no.1
• /
• pp.127-136
• /
• 2022

In this work a multi-fidelity non-intrusive polynomial chaos (MF-NIPC) has been applied to a structural wind engineering problem in architectural design for the first time. In architectural design it is important to design structures that are safe in a range of wind directions and speeds. For this reason, the computational models used to design buildings and bridges must account for the uncertainties associated with the interaction between the structure and wind. In order to use the numerical simulations for the design, the numerical models must be validated by experi-mental data, and uncertainties contained in the experiments should also be taken into account. Uncertainty Quantifi-cation has been increasingly used for CFD simulations to consider such uncertainties. Typically, CFD simulations are computationally expensive, motivating the increased interest in multi-fidelity methods due to their ability to lev-erage limited data sets of high-fidelity data with evaluations of more computationally inexpensive models. Previous-ly, the multi-fidelity framework has been applied to CFD simulations for the purposes of optimization, rather than for the statistical assessment of candidate design. In this paper MF-NIPC method is applied to flow around a rectan-gular 5:1 cylinder, which has been thoroughly investigated for architectural design. The purpose of UQ is validation of numerical simulation results with experimental data, therefore the radius of curvature of the rectangular cylinder corners and the angle of attack are considered to be random variables, which are known to contain uncertainties when wind tunnel tests are carried out. Computational Fluid Dynamics (CFD) simulations are solved by a solver that employs the Finite Element Method (FEM) for two turbulence modeling approaches of the incompressible Navier-Stokes equations: Unsteady Reynolds Averaged Navier Stokes (URANS) and the Large Eddy simulation (LES). The results of the uncertainty analysis with CFD are compared to experimental data in terms of time-averaged pressure coefficients and bulk parameters. In addition, the accuracy and efficiency of the multi-fidelity framework is demonstrated through a comparison with the results of the high-fidelity model.