• Journal of Korea Water Resources Association
• /
• v.31 no.1
• /
• pp.45-57
• /
• 1998

The purpose of this paper is to develop a neural network model in order to forecast flood inflow into the reservoir that has the nature of uncertainty and nonlinearity. The model has the features of multi-layered structure and parallel multi-connections. To develop the model. backpropagation learning algorithm was used with the Momentum and Levenberg-Marquardt techniques. The former technique uses gradient descent method and the later uses gradient descent and Gauss-Newton method respectively to solve the problems of local minima and for the speed of convergency. Used data for learning are continuous fixed real values of input as well as output to emulate the real physical aspects. after learning process. a reservoir inflows forecasting model at flood period was constructed. The data for learning were used to calibrate the developed model and the results were very satisfactory. applicability of the model to the Chungju Mlultipurpose Reservoir proved the availability of the developed model.

• Polymer(Korea)
• /
• v.30 no.2
• /
• pp.97-107
• /
• 2006

In spite of excellent thermomechanical performance the majority of aromatic polyimides are so poor in processability due to their high backbone rigidity that their applications are greatly limited. The introduction of long side chains not only enhances their processiblity but also makes useful contribution to discovering new application fields. In this article, a variety of novel aromatic polyimides with flexible side chains were prepared either from new dimines or new dianhydrides to measure the influence of the side chains on structure and properties of the polymers and their new applications as liquid crystal alignment layers, photosensitive polymers, alternating multilayer nano-films and photoluminescent materials are discussed.

• Korean Journal of Optics and Photonics
• /
• v.27 no.4
• /
• pp.143-150
• /
• 2016

The thickness of each layer in a multilayered system is determined by a Fourier-transform method using spectroscopic reflectance measurements. To verify this method, we first generate theoretical reflectance spectra for three layers, and these are fast-Fourier-transformed using our own Matlab program. Each peak of the Fourier-transformed delta function denotes the optical thickness of each layer, and these are transformed to physical thicknesses. The relative thickness error of the theoretical model is less than 1.0% while a layer's optical thickness is greater than 730 nm. A PI-(thin $SiO_2$)-PImultilayeredstructure produced by the bar-coating method was analyzed, and the thickness errors compared to SEM measurements. Even though this Fourier-transform method requires knowing the film order and the refractive index of each layer prior to analysis, it is a fast and nondestructive method for the analysis of multilayered structures.

• Korean Journal of Food Science and Technology
• /
• v.27 no.4
• /
• pp.525-531
• /
• 1995

Three quality factors of bread during baking process were measured to develop neural network models for bread baking process. Firstly, volume and browning changes during bread baking process were measured using image processing technique and temperature changes inside the bread during process were measured by K-type thermocouples. Relationships among them showed nonlinearity. Secondly, multilayer perception structure with error back propagation learning was used to construct neural network models. Three neural network models for volume, browning, and bread temperature were developed respectively. Developed models showed good performance with predictive error of 4.62% for volume and browning changes after 30 seconds, 7.38% for volume and browning changes after 2 minutes, and 1.09% for temperature change inside the bread respectively.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.42 no.2 s.302
• /
• pp.39-48
• /
• 2005

This paper proposes a novel approach to recognize nine kinds of motion for a multifunction myoelectric hand, acquiring four channel EMG signals from electrodes placed on the forearm. To analyze EMG with properties of nonstationary signal, time-frequency features are extracted by wavelet packet transform. For dimensionality reduction and nonlinear mapping of the features, we also propose a feature projection composed of PCA and SOFM. The dimensionality reduction by PCA simplifies the structure of the classifier, and reduces processing time for the pattern recognition. The nonlinear mapping by SOFM transforms the PCA-reduced features to a new feature space with high class separability. Finally a multilayer neural network is employed as the pattern classifier. From experimental results, we show that the proposed method enhances the recognition accuracy, and makes it possible to implement a real-time pattern recognition.

• Journal of Radiation Protection and Research
• /
• v.41 no.3
• /
• pp.185-190
• /
• 2016

Background: In order to measure neutron energy spectra, the conventional Bonner Sphere Spectrometers (BSS) are widely used. In this spectrometer, several measurements with different size Bonner spheres are required. Operators should, therefore, place these spheres in several times to a measurement point where radiation dose might be relatively high. In order to reduce this effort, novel neutron energy spectrometer using an onion-like single Bonner sphere was proposed in our group. This Bonner sphere has multiple sensitive spherical shell layers in the single sphere. In this spectrometer, a band-shaped thermal neutron detection medium, which consists of a LiF-ZnS mixed powder scintillator sheet and a wavelength-shifting (WLS) fiber readout, was looped to each sphere at equal angular intervals. Amount of LiF neutron converter is reduced near polar region, where the band-shaped detectors are concentrated, in order to uniform the directional sensitivity. The LiF-ZnS mixed powder has an advantage of extremely high light yield. However, since it is opaque, scintillation photons cannot be collect uniformly. This type of detector shows no characteristic shape in the pulse height spectrum. Subsequently, it is difficult to set the pulse height discrimination level. This issue causes sensitivity fluctuation due to gain instability of photodetectors and/or electric modules. Materials and Methods: In order to solve this problem, we propose to replace the LiF-ZnS mixed powder into a flexible and Transparent RUbber SheeT type $LiCaAlF_6$ (TRUST LiCAF) scintillator. TRUST LiCAF scintillator can show a peak shape corresponding to neutron absorption events in the pulse height spectrum. Results and Discussion: We fabricated the prototype detector with five sensitive layers using TRUST LiCAF scintillator and conducted basic experiments to evaluate the directional uniformity of the sensitivity. Conclusion: The fabricated detector shows excellent directional uniformity of the neutron sensitivity.

• Journal of the Korean Magnetics Society
• /
• v.13 no.5
• /
• pp.198-203
• /
• 2003

To investigate whether the use of hard magnetic film is available to generate bias magnetic field for a magnetoimpedance sensor, the magnetic properties of SmCo hard magnetic films were investigated as a function of their compositions. The saturation magnetization decreased with Sm content increasing in SmCo films. And, the coercive force increased in the extent of Sm content of 28 at%, but decreased as Sm content increased moreover. The bias field effect of SmCo film to amorphous CoZrNb film was investigated with the magnetization corves, permeabilities, and magnetic domain structures of SmCo/CoZrNb multilayers. The bias field of about 60 Oe was observed in the film with 3 mm ${\times}$ 0.5 mm, which can be constructed as a MI sensor, and this result strongly indicates that the bias field generated from a hard magnetic film is adequate to enhance the sensitivity of a MI sensor with hard/soft magnetic multilayer structure.

• Journal of the Korean Vacuum Society
• /
• v.19 no.2
• /
• pp.155-160
• /
• 2010

The effect of process temperature of a final annealing step in the fabrication of phase change memory (PCM) devices was investigated. Discrete PCM devices employing $Ge_2Sb_2Te_5$ (GST) films as an active element were made in a pore-style configuration, and they were annealed at various temperatures ranging from 160 to $300^{\circ}C$. The behaviors of cell resistance change from SET resistance to RESET resistance were totally different according to the annealing temperatures. There was a critical annealing temperature for the fabrication of normal PCM devices and abnormal operations were observed in some devices annealed at temperatures lower or higher than the critical temperature. Those influences of annealing temperature seem closely related to the thermal stability of a top electrode/GST/heating layer multilayer structure in the PCM devices.

• Journal of the Korean Ceramic Society
• /
• v.35 no.10
• /
• pp.1040-1048
• /
• 1998

Ferroelectric properties of{{{{ { { { {SrBi }_{2 }Ta }_{2-x }Nb }_{x }O }_{9 } }} (SBIN) thin films were affected by the amount of Bi content in SBTN. The addition of Bi into the SBTN films could be accomplished by heat treating SBTN/Bi2O3/SBTN het-erostructures fabricated by r.f. magnetron sputtering method. The variation of Bi content was controlled by changing the thickness of the sandwiched Bi2O3 in SBTN/Bi2O3/SBTN heterostructure from 50 to 400$\AA$. As changing the thickness of the sandwiched Bi2O3 in Bi2O3 films was increased from 0 to 100$\AA$ the grain grew faster and the ferroelectric pro-perties were improved. On the other hand when the thickness of Bi2O3 films was thicker than 150$\AA$ the fer-roelectric properties were deteriorated Especially for SBTN thin films inserted by 400$\AA$ Bi2O3 layer a Bi2Phase appeared as a second phase resulting in poor ferroelectric properties. The maximum remanent po-larization (2Pr) and coercive field(Ec) were obtained for the SBTN/Bi2O3/(100$\AA$)/SBTN thin films. In this case 2Pr and Ec were 14.75 $\mu$C/cm2 and 53.4kV/cm at an applied voltage of 3V respectively.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2000.02a
• /
• pp.162-162
• /
• 2000

We have studied magnetic properties of Co/Pt multilayered films which have attracted great interest as high-density magneto-optical (MO) recording media due to their good MO properties. For this study, [Pt(45 )/Co(35 )]$\times$8 films were deposited with a Pt buffer layer of 60 on Si(100) substrate by alternating electron-beam evaporation in a high vacuum and were ion beam mixed by using 80keV Ar+ at 25$0^{\circ}C$. Especially, an external magnetic field was added to help changing magnetic property during ion beam mixing (IBM). The intermixing of Co and Pt layers after IBM was confirmed with Rutherford Backscattering Spectroscopy (RBS) and Transmission Electron Microscopy (TEM). The MO property of the film was measured with magneto-optical Kerr spectrometer and the change of magnetic easy axis in the film plane was observed from Ker loop data. This anomalous result might be correlated with the change of atomic structure due to the intermixing effect.