• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.33 no.6
• /
• pp.288-293
• /
• 2023

To optimize the process parameters of laser powder bed fusion process to fabricate the high density STS316L alloy, the effect of laser power, scanning speed and hatching distance on the relative density was studied. Tensile properties of additively manufactured STS316L alloy using optimized parameters was also evaluated according to the build direction. As a result of additive manufacturing process under the energy density of 55.6 J/mm³, 83.3 J/mm³ and 111.1 J/mm³, high density STS316L specimens was suitably fabricated when the energy density, power and scan speed were 83.3 J/mm³, 225 W and 1000 mm/s, respectively. The yield strength, ultimate tensile strength, and elongation of STS316L specimens in direction perpendicular to the build direction, show the most competitive values. Anisotropic shape of the pores and the lack of fusion defects probably caused strain localization which result in deterioration of tensile properties.

• Journal of the Korean institute of surface engineering
• /
• v.56 no.6
• /
• pp.412-419
• /
• 2023

Anion exchange membrane electrolysis is considered a promising next-generation hydrogen production technology that can produce low-cost, clean hydrogen. However, anion exchange membrane electrolysis technology is in its early stages of development and requires intensive research on electrodes, which are a key component of the catalyst-system interface. In this study, we optimized the pressure conditions of the hot-pressing process to manufacture cobalt oxide electrodes for the development of a high uniformity and high adhesion electrode production process for the oxygen evolution reaction. As the pressure increased, the reduction of pores within the electrode and increased densification of catalytic particles led to the formation of a uniform electrode surface. The cobalt oxide electrode optimized for pressure conditions exhibited improved catalytic activity and durability. The optimized electrode was used as the anode in an AEMWE single cell, exhibiting a current density of 1.53 A cm^-2 at a cell voltage of 1.85 V. In a durability test conducted for 100 h at a constant current density of 500 mA cm^-2, it demonstrated excellent durability with a low degradation rate of 15.9 mV kh^-1, maintaining 99% of its initial performance.

• Resources Recycling
• /
• v.33 no.1
• /
• pp.22-30
• /
• 2024

To synthesize the Ti₃AlC₂ MAX phase, a crucial precursor for generating the two-dimensional material MXene, the use of Ti scrap as an initial material is an economically feasible approach. This study aims to optimize the synthesis conditions for the phase fraction of the Ti₃AlC₂ MAX phase utilizing Ti scrap as the Ti source. The deoxidation of Ti powders, prepared through the hydrogenation-dehydrogenation process from Ti scrap, was effectively accomplished using the deoxidation in solid-state (DOSS) process. The optimal synthesis conditions were established by blending DOSS-Ti, Al, and graphite powders with particle sizes ranging from 25 ~ 32 ㎛ in a molar ratio of 3:1.1:2. The resulting phase fractions were as follows: Ti₃AlC₂ at 97.25 wt.%, TiC at 0.93 wt.%, and Al₃Ti at 1.82 wt.%. Furthermore, the oxygen content of the Ti₃AlC₂ MAX powder, spanning from 25 ~ 45 ㎛, was measured at 4,210 ppm.

• Information Systems Review
• /
• v.24 no.1
• /
• pp.39-57
• /
• 2022

This paper proposes the GMOPTBoost algorithm to improve the performance of the AdaBoost algorithm for bankruptcy prediction in which class imbalance problem is inherent. AdaBoost algorithm has the advantage of providing a robust learning opportunity for misclassified samples. However, there is a limitation in addressing class imbalance problem because the concept of arithmetic mean accuracy is embedded in AdaBoost algorithm. GMOPTBoost can optimize the geometric mean accuracy and effectively solve the category imbalance problem by applying Gaussian gradient descent. The samples are constructed according to the following two phases. First, five class imbalance datasets are constructed to verify the effect of the class imbalance problem on the performance of the prediction model and the performance improvement effect of GMOPTBoost. Second, class balanced data are constituted through data sampling techniques to verify the performance improvement effect of GMOPTBoost. The main results of 30 times of cross-validation analyzes are as follows. First, the class imbalance problem degrades the performance of ensembles. Second, GMOPTBoost contributes to performance improvements of AdaBoost ensembles trained on imbalanced datasets. Third, Data sampling techniques have a positive impact on performance improvement. Finally, GMOPTBoost contributes to significant performance improvement of AdaBoost ensembles trained on balanced datasets.

• Animal Bioscience
• /
• v.37 no.5
• /
• pp.795-806
• /
• 2024

Objective: Karan Fries (KF), a high-producing composite cattle was developed through crossing indicine Tharparkar cows with taurine bulls (Holstein Friesian, Brown Swiss, and Jersey), to increase the milk yield across India. This composite cattle population must maintain sufficient genetic diversity for long-term development and breed improvement in the coming years. The level of linkage disequilibrium (LD) measures the influence of population genetic forces on the genomic structure and provides insights into the evolutionary history of populations, while the decay of LD is important in understanding the limits of genome-wide association studies for a population. Effective population size (N_e) which is genomically based on LD accumulated over the course of previous generations, is a valuable tool for e valuation of the genetic diversity and level of inbreeding. The present study was undertaken to understand KF population dynamics through the estimation of N_e and LD for the long-term sustainability of these breeds. Methods: The present study included 96 KF samples genotyped using Illumina HDBovine array to estimate the effective population and examine the LD pattern. The genotype data were also obtained for other crossbreds (Santa Gertrudis, Brangus, and Beefmaster) and Holstein Friesian cattle for comparison purposes. Results: The average LD between single nucleotide polymorphisms (SNPs) was r² = 0.13 in the present study. LD decay (r² = 0.2) was observed at 40 kb inter-marker distance, indicating a panel with 62,765 SNPs was sufficient for genomic breeding value estimation in KF cattle. The pedigree-based N_e of KF was determined to be 78, while the N_e estimates obtained using LD-based methods were 52 (SNeP) and 219 (genetic optimization for N_e estimation), respectively. Conclusion: KF cattle have an N_e exceeding the FAO's minimum recommended level of 50, which was desirable. The study also revealed significant population dynamics of KF cattle and increased our understanding of devising suitable breeding strategies for long-term sustainable development.

• Information Systems Review
• /
• v.24 no.4
• /
• pp.23-40
• /
• 2022

Although Support Vector Machine(SVM) has been used in various fields such as bankruptcy prediction model, the hyperplane learned by SVM in class imbalance problem can be severely skewed toward minority class and has a negative impact on performance because the area of majority class is expanded while the area of minority class is invaded. This study proposed optimized uneven margin SVM(OPT-UMSVM) combining threshold moving or post scaling method with UMSVM to cope with the limitation of the traditional even margin SVM(EMSVM) in class imbalance problem. OPT-UMSVM readjusted the skewed hyperplane to the majority class and had better generation ability than EMSVM improving the sensitivity of minority class and calculating the optimized performance. To validate OPT-UMSVM, 10-fold cross validations were performed on five sub-datasets with different imbalance ratio values. Empirical results showed two main findings. First, UMSVM had a weak effect on improving the performance of EMSVM in balanced datasets, but it greatly outperformed EMSVM in severely imbalanced datasets. Second, compared to EMSVM and conventional UMSVM, OPT-UMSVM had better performance in both balanced and imbalanced datasets and showed a significant difference performance especially in severely imbalanced datasets.

• Advanced Industrial SCIence
• /
• v.3 no.2
• /
• pp.31-37
• /
• 2024

This paper presents a single and 2-patch microstrip array antenna operated on a frequency of 10.3GHz(x-band). It outlines the process of designing a microstrip patch array antenna using CST MWS. Initially, a single microstrip antenna was designed, followed by optimization using CST MWS to attain optimal return losses and gain. Subsequently, the design was expanded to create a 2×1 microstrip inset-fed array antenna for the X-band applications. The construction material is Roger RO4350B, with specific dimensions (h=0.79mm, 𝜖_r = 3.54). The achieved results include an S11 of -18dB at the resonant frequency (10.3GHz), a gain of 9.82dBi, a bandwidth of 0.165GHz, and a 3-dB beamwidth of 30°, 121° in Az(𝜑=0) and El(𝜑=90) plane, respectively. The future plan involves the fabrication of this array antenna and further expansion to a 4×4 array of microstrip antennas. It is then incorporated on the X-band applications for practical uses.

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.11 no.2
• /
• pp.82-91
• /
• 2006

We investigated the possibility of using effluents from a municipal sewage treatment plant (STP) as tracers a tracer for hydrologic studies of rivers. The possibility was checked in a 12-km long reach downstream of Jeonju Municipal Sewage Treatment Plant (JSTP). Time-series monitoring of the water chemistry reveals that chemical compositions of the effluent from the JSTP are fluctuating within a relatively wide range during the sampling period. In addition, the signals from the plant were observed at the downstream stations consecutively with increasing time lags, especially in concentrations of the conservative chemical parameters (concentrations f3r chloride and sulfate, total concentration of major cations, and electric conductivity). Based on this observation, we could estimate the stream flow (Q), velocity (v), and dispersion coefficient (D). A 1-D nonreactive solute-transport model with automated optimization schemes was used for this study. The values of Q, v, and D estimated from this study varied from 6.4 to $9.0m^3/sec$ (at the downstream end of the reach), from 0.06 to 0.10 m/sec, and from 0.7 to $6.4m^2/sec$, respectively. The results show that the effluent from a large-scaled municipal STP frequently provides good, multiple natural tracers far hydrologic studies.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.37 no.2
• /
• pp.85-93
• /
• 2024

In this paper, we propose a method for establishing a state-space equation model for the motion analysis of floating structures subjected to wave loads, by applying system-identification techniques. Traditionally, the motion of floating structures has been analyzed in the time domain by integrating the Cummins equation over time, which utilizes a convolution integral term to account for the effects of the retardation function. State-space equation models have been studied as a way to efficiently solve floating-motion equations in the time domain. The proposed approach outlines a procedure to derive the target transfer function for the load-displacement input/output relationship in the frequency domain and subsequently determine the state-space equation that closely approximates it. To obtain the state-space equation, the method employs the N4SID system-identification method and an optimization approach that treats the coefficients of the numerator and denominator polynomials as design variables. To illustrate the effectiveness of the proposed method, we applied it to the analysis of a single-degree-of-freedom model and the motion of a six-degree-of-freedom barge. Our findings demonstrate that the presented state-space equation model aligns well with the existing analysis results in both the frequency and time domains. Notably, the method ensures computational accuracy in the time-domain analysis while significantly reducing the calculation time.

• Journal of Internet of Things and Convergence
• /
• v.10 no.2
• /
• pp.1-15
• /
• 2024

The construction of smart communities is a new method and important measure to ensure the security of residential areas. In order to solve the problem of low accuracy in face recognition caused by distorting facial features due to monitoring camera angles and other external factors, this paper proposes the following optimization strategies in designing a face recognition network: firstly, a global graph convolution module is designed to encode facial features as graph nodes, and a multi-scale feature enhancement residual module is designed to extract facial keypoint features in conjunction with the global graph convolution module. Secondly, after obtaining facial keypoints, they are constructed as a directed graph structure, and graph attention mechanisms are used to enhance the representation power of graph features. Finally, tensor computations are performed on the graph features of two faces, and the aggregated features are extracted and discriminated by a fully connected layer to determine whether the individuals' identities are the same. Through various experimental tests, the network designed in this paper achieves an AUC index of 85.65% for facial keypoint localization on the 300W public dataset and 88.92% on a self-built dataset. In terms of face recognition accuracy, the proposed network achieves an accuracy of 83.41% on the IBUG public dataset and 96.74% on a self-built dataset. Experimental results demonstrate that the network designed in this paper exhibits high detection and recognition accuracy for faces in surveillance videos.