• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.53 no.1
• /
• pp.40-49
• /
• 2004

In this paper, we propose a new architecture of Genetic Algorithms(GAs)-based Self-Organizing Polynomial Neural Networks(SOPNN), discuss a comprehensive design methodology and carry out a series of numeric experiments. The conventional SOPNN is based on the extended Group Method of Data Handling(GMDH) method and utilized the polynomial order (viz. linear, quadratic, and modified quadratic) as well as the number of node inputs fixed (selected in advance by designer) at Polynomial Neurons (or nodes) located in each layer through a growth process of the network. Moreover it does not guarantee that the SOPNN generated through learning has the optimal network architecture. But the proposed GA-based SOPNN enable the architecture to be a structurally more optimized network, and to be much more flexible and preferable neural network than the conventional SOPNN. In order to generate the structurally optimized SOPNN, GA-based design procedure at each stage (layer) of SOPNN leads to the selection of preferred nodes (or PNs) with optimal parameters- such as the number of input variables, input variables, and the order of the polynomial-available within SOPNN. An aggregate performance index with a weighting factor is proposed in order to achieve a sound balance between approximation and generalization (predictive) abilities of the model. A detailed design procedure is discussed in detail. To evaluate the performance of the GA-based SOPNN, the model is experimented with using two time series data (gas furnace and NOx emission process data of gas turbine power plant). A comparative analysis shows that the proposed GA-based SOPNN is model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• Journal of Environmental Impact Assessment
• /
• v.23 no.2
• /
• pp.101-111
• /
• 2014

The research was carried out in order to find climate factors which determine the distribution of Neolitsea sericea, and the potential habitats (PHs) under the current climate and three climate change scenario by using species distribution models (SDMs). Four climate factors; the minimum temperature of the coldest month (TMC), the warmth index (WI), summer precipitation (PRS), and winter precipition (PRW) : were used as independent variables for the model. Three general circulation models under A1B emission scenarios were used as future climate scenarios for the 2050s (2040~2069) and 2080s (2070~2099). Highly accurate SDMs were obtained for N. sericea. The model of distribution for N. sericea constructed by SDMs showed that minimum temperature of the coldest month (TMC) is a major climate factor in determining the distribution of N. sericea. The area above the $-4.4^{\circ}C$ of TMC revealed high occurrence probability of the N. sericea. Future PHs for N. sericea were projected to increase respectively by 4 times, 6.4 times of current PHs under 2050s and 2080s. It is expected that the potential of N. sericea habitats is expanded gradually. N. sericea is applicable as indicator species for monitoring in the Korean Peninsula. N. sericea is necessary to be monitored of potential habitats.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.3 no.1
• /
• pp.37-53
• /
• 1993

The polynuclear hydrocarbons (PAHs) emitted from coke oven standpipe were sampled using three sampling systems, including glass fiber filter+silver membrane filter, glass fiber filter+silver membrane filter+XAD-2 adsorbent tube, PTFE membrane filter+XAD-2 adsorbent tube, extracted by methylene chloride and analysed by gas chromathography using flame ionization detector. The results of this study were as follows. 1. Because the amounts of coke oven emissions(COE) were large, the analyses of PAHs were simple and possible without evaporation and concentration. Although the generation of COE was high during early stage of coking, the airborne concentration of PAHs was low and increased during late coking. 2. The contents of PAHs in COE were 1.35-2.81%. 3. The index components of PAHs were fluoranthene and pyrene. Their correlation coefficient to total PAHs were 0.96, 0.95, respectively. 4. The particulate PAHs were sampled by filter and gaseous PAHs by adsorbent tube. The collection efficiency of glass fiber filter+silver membrane filter was 20% of total amount sampled by filters+adsorbent and PTFE membrane filter 50%. Adsorbent tube must be attached to the filter to collect light and small PAH components. 5. The generation of acenaphthene and indeno (1,2,3-cd) pyrene were low and concentrations of fluorene and anthracene were $20-40ug/m^3$ throughout coking time. Other PAH eoncentrations were sometimes high. The generation of PAHs was low at 4-6 hours of coking time. The gaseous PAHs were generated earlier than particulate PAHs.

• Geomechanics and Engineering
• /
• v.11 no.3
• /
• pp.345-359
• /
• 2016

In order to investigate the influence of the interfacial angel on failure characteristics and mechanism of combined coal-rock mass, 35 uniaxial/biaxial compressive simulation tests with 5 different interfacial angels of combined coal-rock samples were conducted by PFC2D software. The following conclusions are drawn: (1) The compressive strength and cohesion decrease with the increase of interfacial angle, which is defined as the angle between structure plane and the exterior normal of maximum principal plane, while the changes of elastic modulus and internal friction angle are not obvious; (2) The impact energy index $K_E$ decreases with the increase of interfacial angle, and the slip failure of the interface can be predicted based on whether the number of acoustic emission (AE) hits has multiple peaks or not; (3) There are four typical failure patterns for combined coal-rock samples including I (V-shaped shear failure of coal), II (single-fracture shear failure of coal), III (shear failure of rock and coal), and IV (slip rupture of interface); and (4) A positive correlation between interfacial angle and interface effect is shown obviously, and the interfacial angle can be divided into weak-influencing scope ($0-15^{\circ}$), moderate-influencing scope ($15-45^{\circ}$), and strong-influencing scope (> $45^{\circ}$), respectively. However, the confining pressure has a certain constraint effect on the interface effect.

• Bulletin of the Korean Space Science Society
• /
• 1992.10a
• /
• pp.21-21
• /
• 1992

The X-ray binary pulsar Her X-1 was observed with Ginga on 1988 August 28 during the orbital phase of 0.76 to 0.85 at the main-on phase of the 35 day cycle. During the observations the X-ray intensity varied by a factor of five or more on a time scale as short as 30 sec, due mostly to the soft X-ray absorption in the pre-eclipse phase. From the studies of pulse profiles and energy spectra, we revealed that there exists in the dipphase an unpulsed component which is "3% of the intensity at the non-absorbed high-level. We suggest that scattering of the source continuum by the optically thin hot corona is responsible for the unpoised component. In the spectral analysis, we find that the high-state non-absorbed spectra can be fitted by a power-law without absorption, and the spectra observed in the different abgorption states by two components of a power-law with the same photon index. An iron-K emission line is required in to the cases of fitting. The estimated equivalent width of the iron line varies from 0.18 to 0.51 key according to the change in the absorption column density along the line of sight. We suggest that the fluorescent iron line arises in a cool and relatively small region, like the Alfvensur face, and may be partially intercepted by the optically thick gas cloud passing across the line of sight.1 Korea Astronomy Observatorya The Institute of Space and Astronautical Science in Japan3 Korea Advanced Institute of Science and Technology

• Journal of Korean Society on Water Environment
• /
• v.32 no.1
• /
• pp.98-107
• /
• 2016

This study examined the effects of environmental conditions and the presence of refractory organic matter on oxidation rates of total organic carbon (TOC) measurements based on high temperature combustion and ultraviolet-sulfate methods. Spectroscopic indices for prediction of oxidation rates were also explored using the UV spectra and fluorescence excitation-emission matrix (EEM) of humic acids. Furthermore, optimum TOC instrument conditions were suggested by comparing oxidation rates of a standard TOC material under various conditions. Environmental conditions included salts, reduced ions, and suspended solids. Salts had the greatest influence on oxidation rates in the UV-sulfate method. However, no effect was detected in the high temperature combustion method. The UV-sulfate method showed lower humic substance oxidation rates, refractory natural organic matter, compared to the other methods. TOC oxidation rates for the UV-sulfate method were negatively correlated with higher specific-UV absorbance, humification index, and humic-like EEM peak intensities, suggesting that these spectroscopic indices could be used to predict TOC oxidation rates. TOC signals from instruments using the UV-sulfate method increased with increasing chamber temperature and increasing UV exposure durations. Signals were more sensitive to the former condition, suggesting that chamber temperature is important for improving the TOC oxidation rates of refractory organic matter.

• Radiation Oncology Journal
• /
• v.33 no.1
• /
• pp.50-56
• /
• 2015

Purpose: To investigate the coverage of axillary lymph node with tangential breast irradiation fields by using virtual lymph node (LN) analysis. Materials and Methods: Forty-eight women who were treated with whole breast irradiation after breast-conserving surgery were analyzed. The axillary and breast volumes were delineated according to the Radiation Therapy Oncology Group (RTOG) contouring atlas. To generate virtual LN contours, preoperative fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) scans with identifiable LN were fused with the CT scans, and the virtual LN contour were delineated on the CT. Results: The median level I and II axillary volume coverage percentages at the $V_{D95%}$ line were 33.5% (range, 5.3% to 90.4%) and 0.6% (range, 0.0% to 14.6%), respectively. Thirty-one LNs in 18 patients were delineated (26 in level I and 5 in level II). In the level I axilla, 84.6% of virtual LNs were encompassed by the 95% isodose line. In the level II axilla, by contrast, none of the virtual LNs were encompassed by the 95% isodose volumes. There was a substantial discrepancy between the RTOG contouring atlas-based axillary volume analysis and the virtual LN analysis, especially for the level I axillary coverage. The axillary volume coverage was associated with the body mass index (BMI) and breast volume. Conclusion: The tangential breast irradiation did not deliver adequate therapeutic doses to the axillary region, particularly those in the level II axilla. Patients with small breast volumes or lower BMI showed reduced axillary coverage from the tangential breast fields. For axillary LN irradiation, individualized anatomy-based radiation fields for patients would be necessary.

• Journal of the Korean Ceramic Society
• /
• v.54 no.1
• /
• pp.49-54
• /
• 2017

In this study, based on hydrolysis and condensation via $St{\ddot{o}}ber$ process of sol-gel method, synthesis of mono-dispersed silica nanoparticles was carried out with hydrophilic solvent. This operation was expected to be a more simplified process than that with organic solvent. Based on the sol-gel method, which involves simply controlling the particle size, the particle size of the synthesized silica specimens were ranged from 30 to 300 nm by controlling the composition of tetraethylorthosilicate (TEOS), DI water and ammonia solution, and by varying the stirring speeds while maintaining a fixed amount of ethanol. Increasing the content of DI water and decreasing the content of ammonia caused the particle size to decrease, while controlling the stirring speed at a high level of RPMs enabled a decrease of the particle size. X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FT-IR) were utilized to investigate the success factors for synthesizing process; Field emission scanning electron microscopy (FE-SEM) was used to study the effects of the size and morphology of the synthesized particles. To analyze the dispersion properties, zeta potential and particle size distribution (PSD) analyses were utilized.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.30 no.7
• /
• pp.427-431
• /
• 2017

In this paper, we discuss ${\beta}-Ga_2O_3$ thin films that have been grown on freestanding GaN (FS-GaN) using furnace oxidation. A GaN template was grown by horizontalhydride vapor phase epitaxy (HVPE), and FS-GaN was fabricated using the laser lift off (LLO) system. To obtain ${\beta}-Ga_2O_3$ thin film, FS-GaN was oxidized at $900{\sim}1,100^{\circ}C$. Surface and cross-section of prepared ${\beta}-Ga_2O_3$ thin films were observed by field emission scanning electron microscopy (FE-SEM). The single crystal FS-GaNs were changed to poly-crystal ${\beta}-Ga_2O_3$. The oxidized ${\beta}-Ga_2O_3$ thin film at $1,100^{\circ}C$ was peel off from FS-GaN. Next, oxidation of FS-GaNwas investigated for 0.5~12 hours with variation of the oxidation time. The thicknesses of ${\beta}-Ga_2O_3$ thin films were measured from 100 nm to 1,200 nm. Moreover, the 2-theta XRD result indicated that (-201), (-402), and (-603) peaks were confirmed. The intensity of peaks was increased with increased oxidation time. The ${\beta}-Ga_2O_3$ thin film was generated to oxidize FS-GaN.

• International Journal of Control, Automation, and Systems
• /
• v.1 no.2
• /
• pp.194-202
• /
• 2003

In this paper, we introduce a category of Multi-FNN (Fuzzy-Neural Networks) models, analyze the underlying architectures and propose a comprehensive identification framework. The proposed Multi-FNNs dwell on a concept of fuzzy rule-based FNNs based on HCM clustering and evolutionary fuzzy granulation, and exploit linear inference being treated as a generic inference mechanism. By this nature, this FNN model is geared toward capturing relationships between information granules known as fuzzy sets. The form of the information granules themselves (in particular their distribution and a type of membership function) becomes an important design feature of the FNN model contributing to its structural as well as parametric optimization. The identification environment uses clustering techniques (Hard C - Means, HCM) and exploits genetic optimization as a vehicle of global optimization. The global optimization is augmented by more refined gradient-based learning mechanisms such as standard back-propagation. The HCM algorithm, whose role is to carry out preprocessing of the process data for system modeling, is utilized to determine the structure of Multi-FNNs. The detailed parameters of the Multi-FNN (such as apexes of membership functions, learning rates and momentum coefficients) are adjusted using genetic algorithms. An aggregate performance index with a weighting factor is proposed in order to achieve a sound balance between approximation and generalization (predictive) abilities of the model. To evaluate the performance of the proposed model, two numeric data sets are experimented with. One is the numerical data coming from a description of a certain nonlinear function and the other is NOx emission process data from a gas turbine power plant.