• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.5
• /
• pp.457-461
• /
• 2014

Using $0.18{\mu}m$ CMOS process silicon neuron circuit of the pulse type for modeling biological neurons, were designed in the semiconductor integrated circuit. Neuron circuiSt providing is formed by MOS switch for initializing the input terminal of the capacitor to the input current signal, a pulse signal and an amplifier stage for generating an output voltage signal. Synapse circuit that can convert the current signal output of the input voltage signal, using a bump circuit consisting of NMOS transistors and PMOS few. Configure a chain of neurons for verification of the neuron model that provides synaptic neurons and two are connected in series, were performed SPICE simulation. Result of simulation, it was confirmed the normal operation of the synaptic transmission characteristics of the signal generation of nerve cells.

• Bulletin of the Society of Naval Architects of Korea
• /
• v.14 no.1
• /
• pp.1-10
• /
• 1977

The major objects of this report are to supplement data of natural frequencies of thin elastic rectangular plates to the available data, and to give an experimental verification for natural frequencies obtained by Rayleigh-Ritz method, the generation set of which are eigenfunctions of Euler beams. For the first object the following five models, for which data only for the fundamental mode or data only for square plates are available, are adopted; (1) two opposed edges are clamped and the other two opposed edges simply supported (C-C, S-S), (2) one edge is simply supported and the other three edges clamped (C-C, C-S), (3) one edge is free and the other three edges clamped (C-C, C-F), (4) two adjacent edges are clamped and the other two adjacent edges free (C-F, C-F). For the (C-C, S-S) model the frequency equation obtained with the mode shapes assumed as of a single trigonometric series is solved. And for the other four models Rayleigh-Ritz method taking eigenfunctions of Euler beams as the generating set is applied. The numerical examples are obtained up to the fourth, the fifth or the sixth order depending on the range of the aspect ratio (0.1-10.0). The number of terms in the generating set for Rayleigh-Ritz method is fifteen for all models. For the experiment three models made of 3.2mm thickness mild steel plate for general structure use were prepared in following size; $300mm{\times}600mm,\;600mm{\times}600mm\;and\;900mm{\times}600mm$. Their boundary conditions are made to fit (C-C, C-F) condition. From the experiment mechanical impedance curves based on the frequency response method were obtained together with phase relation diagrams. The experimental data are resulted in good conformity to calculated values.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.16 no.3
• /
• pp.59-72
• /
• 2017

As the part of study which handles the measure to use the individual vehicle information of taxi GPS data on signal controls in order to overcome the limitation of Loop detector-based collecting methods of real-time signal control system, this paper conducted series of evaluations and improvements on link travel time, queue vehicle time estimates and traffic condition decision algorithm from the research introduced in 2016. considering the control group and the other, the link travel time has enhanced the travel time and the length of queue vehicle has enhanced the estimated model taking account of the traffic situation. It is analyzed that the accuracy of the average link travel time and the length of queue vehicle are respectably both approximately 95 % and 85%. The traffic condition decision algorithm reflected the improved travel speed and vehicle length. Smoothing was performed to determine the trend of the traffic situation and reduce the fluctuation of the data, and the algorithms have refined so as to reflect the pass period on overflow judgment criterion.

• Proceedings of the Korea Association of Information Systems Conference
• /
• 2000.11a
• /
• pp.1.2-24
• /
• 2000

The field of neural networks has been a recent surge in activity as a result of progress in developments of efficient training algorithms. For this reason, and coupled with the widespread availability of powerful personal computer hardware for running simulations of networks, there is increasing focus on the potential benefits this field can offer. The neural network may be viewed as an advanced pattern recognition technique and can be applied in many areas such as financial time series forecasting, medical diagnostic expert system and etc.. The intention of this study is to build and implement one simple artificial neural networks hereinafter called ANN. For this purpose, some literature survey was undertaken to understand the structures and algorithms of ANN theoretically. Based on the review of theories about ANN, the system adopted 3-layer back propagation algorithms as its learning algorithm to simulate one case of medical diagnostic model. The adopted ANN algorithm was performed in PC by using turbo PASCAL and many input parameters such as the numbers of layers, the numbers of nodes, the number of cycles for learning, learning rate and momentum term. The system output more or less successful results which nearly agree with goals we assumed. However, the system has some limitations such as the simplicity of the programming structure and the range of parameters it can dealing with. But, this study is useful for understanding general algorithms and applications of ANN system and can be expanded for further refinement for more complex ANN algorithms.

• Journal of Welding and Joining
• /
• v.34 no.6
• /
• pp.42-46
• /
• 2016

Hydrogen assisted cracking (HAC) is one of the most complicated problem in welding. Huge amount of studies have been done for decades. Based on them, various standards have been established to avoid HAC. But it is still a chronic problem in industrial field. It is well known that the main causes of the hydrogen crack are residual stress, crack susceptible micro structures and a certain critical level of hydrogen concentration. Even though the exact generating mechanism is unclear till today, it has been reported that the hydrogen level in the weld metal should be managed less than a certain amount to prevent it. Matsuda studied that the residual hydrogen level in the weld metal can be varied even if the initial hydrogen content is same. It is also insisted in this report that the residual hydrogen concentration is in stronger correlation with hydrogen crack than the initial hydrogen content. But, in practical point of view, the residual hydrogen is still hard to consider because measuring hydrogen level is time and cost consuming process. In this regard, numerical analysis is the only solution for considering the residual hydrogen content. Meanwhile, Takahashi showed the possibility of predicting the residual hydrogen by a rigorous FE analysis. But, few commercial software suitable for solving the weld metal hydrogen has been reported yet. In this study, two dimensional thermal - hydrogen coupled analysis was developed by using the commercial FE software MARC. Since the governing equation of the hydrogen diffusion is similar to the heat transfer, it is shown that the heat transfer FE analysis in association with hydrogen diffusion property can be used for hydrogen diffusion analysis. A series of simulation was performed to verify the accuracy of the model. For BOP (Bead-On-Plate) and the multi-pass butt welding simulations, remaining hydrogen contents in the weld metal is well matched with measurements which are referred from Kim and Masamitsu.

• Journal of the Korea Concrete Institute
• /
• v.23 no.6
• /
• pp.765-772
• /
• 2011

The present study summarizes a series of compressive tests on concrete cylinder in order to examine the stressstrain relationship of alkali-activated (AA) slag concrete. The compressive strength and unit weight of concrete tested ranged from 8.6 MPa to 42.2 MPa and from $2,186kg/m^3$ to $2,343kg/m^3$, respectively. A mathematical equation representing the complete stress-strain curve was developed based on test results recorded from 34 concrete specimens. The modulus of elasticity, strain at peak stress, slopes of ascending and descending branches of stress-strain curves were generalized as a function of compressive strength and unit weight of concrete. The mean and standard deviation of the coefficient of variance between measured and predicted curves were 6.9% and 2.6%, respectively. This indicates that the stress-strain relationship of AA slag concrete is represented properly with more accuracy in the proposed model than in some other available models for ordinary portland cement (OPC) concrete.

• Journal of Radiation Protection and Research
• /
• v.41 no.4
• /
• pp.368-372
• /
• 2016

Background: Methodologies for a series of radiological consequence assessments show a distinctive difference according to the design principles of the original nuclear suppliers and their technical standards to be imposed. This is due to the uncertainties of the accidental source term, radionuclide behavior in the environment, and subsequent radiological dose. Both types of PWR and PHWR are operated in Korea. However, technical standards for evaluating atmospheric dispersion have been enacted based on the U.S. NRC's positions regardless of the reactor types. For this reason, it might cause a controversy between the licensor and licensee of a nuclear power plant. Materials and Methods: It was modelled under the framework of the NRC Regulatory Guide 1.145 for light-water reactors, reflecting the features of heavy-water reactors as specified in the Canadian National Standard and the modelling features in MACCS2, such as atmospheric diffusion coefficient, ground deposition, surface roughness, radioactive plume depletion, and exposure from ground deposition. Results and Discussion: An integrated accident consequence assessment code, ACCESS (Accident Consequence Assessment Code for Evaluating Site Suitability), was developed by taking into account the unique regulatory positions for reactor types under the framework of the current Korean technical standards. Field tracer experiments and hand calculations have been carried out for validation and verification of the models. Conclusion: The modelling approaches of ACCESS and its features are introduced, and its applicative results for a hypothetical accidental scenario are comprehensively discussed. In an applicative study, the predicted results by the light-water reactor assessment model were higher than those by other models in terms of total doses.

• Phonetics and Speech Sciences
• /
• v.12 no.3
• /
• pp.55-63
• /
• 2020

Voice conversion can be applied to various voice processing applications. It can also play an important role in data augmentation for speech recognition. The conventional method uses the architecture of voice conversion with speech synthesis, with Mel filter bank as the main parameter. Mel filter bank is well-suited for quick computation of neural networks but cannot be converted into a high-quality waveform without the aid of a vocoder. Further, it is not effective in terms of obtaining data for speech recognition. In this paper, we focus on performing voice-to-voice conversion using only the raw spectrum. We propose a deep learning model based on the transformer network, which quickly learns the voice conversion properties using an attention mechanism between source and target spectral components. The experiments were performed on TIDIGITS data, a series of numbers spoken by an English speaker. The conversion voices were evaluated for naturalness and similarity using mean opinion score (MOS) obtained from 30 participants. Our final results yielded 3.52±0.22 for naturalness and 3.89±0.19 for similarity.

• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.9 no.3
• /
• pp.334-338
• /
• 1999

Porous silicon layers were fabricated with various conditions of HF concentration and current density. And their masses were measured. From these data, the porosity and fractal dimension were estimated and analyzed. We found that the porosity was proportional to the current density when the anodic reaction time was fixed and the constant values of fractal dimension could be estimated from a series of data with fixed HF concentration. The values of fractal dimension were decreased with increasing HF concentration. The obtained porosity and fractal dimension were compared with the 2-dimensional computer simulation based on diffusion limited deposition model. According to the simulation, the porosity was proportional to the diffusion length and the fractal dimension was inversely proportional to the diffusion length. Since, the diffusion length is proportional to current density and inversely proportional to base concentration, our experimental data qualitatively agreed with the results from the simulation. The porosity obtained by experiments, however, was not consistent with the results by simulation.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.25 no.5
• /
• pp.376-384
• /
• 2020

An energy storage system is composed of lithium-ion batteries in modern applications. Batteries are regarded as storage devices for renewable and residual energy. The failure of batteries can cause the performance reduction and explosion of battery systems. High maintenance cost is essential when dealing with the problem of battery safety. Therefore an accurate health diagnosis is required to ensure the high reliability of battery systems. A battery pack is a combination of single cells in series and parallel connections. A battery pack has to consider various factors to assess battery health. Battery health involves conventional factors and additional factors, such as cell-to-cell imbalance. For large applications, state-of-health (SOH) can be inaccurate because of the lack of factors that indicate the state of the battery pack. In this study, six characterization factors are proposed for improving the SOH estimation of battery packs. The six proposed characterization factors can be regarded as health indicators (HIs). The six HIs are applied to the principal component analysis (PCA) algorithm. To reflect information regarding capacity, voltage, and temperature, the PCA algorithm extracts new degradation factors by using the six HIs. The new degradation factors are applied to a multiple regression model. Results show the advancement and improvement of SOH estimation.