• Proceedings of the KOSOMBE Conference
• /
• v.1992 no.11
• /
• pp.47-50
• /
• 1992

This paper describes pattern classification algorithm of ECG using back-propagation neural network. We presents new feature extractor using second order approximating function as the input signals of neural network. We use 9 significant parameters which were extracted by feature extractor. 5 most characterized ECG signal pattern is classified accurately by neural network. We use AHA database to evaluate the performance ol the proposed pattern classification algorithm.

• Tribology and Lubricants
• /
• v.10 no.1
• /
• pp.69-77
• /
• 1994

Oil flow pass of hermetic scroll compressor with back pressure chamber was described. Dynamic analysis was preceded in order to obtain the loads on the lubricating contacts. The mobility method of dynamically loaded journal bearings was applied to the crank jornal bearing and lower main bearing, and they could be designed to operate under fluid film lubrication. From the consideration of their film thicknesses and oil flow rates, optimal bearing clearances or other bearing dimensions could be assessed. The major friction loss was calculated to be from the axial force between the two scrolls. Therefore, it was suggested that the designers should be careful to reduce the over-turning moment on the orbiting scroll.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2005.10a
• /
• pp.402-405
• /
• 2005

Such press-forming process are the used machine ability and the characteristic, used material, tile accuracy of the part, condition of a process are considered the designed. In order to estimate in automotive sheet forming processes used AutoForm software. A through in simulation result comparison with experimentation result, it was possible to know that much the same estimated spring-back through a forming analysis. By making apply this to an industrial site the productivity improvement and cost reduction etc. effect able was predicted.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.13 no.2
• /
• pp.16-21
• /
• 2005

The performance of reciprocating internal combustion engine is a strong function of the air intake system configuration. In order to improve engine power, it is essential to optimize the air induction system. In this study, a numerical investigation has been carried out for the three-dimensional flow and pressure characteristics in air intake system of a large-sized commercial bus. CFD simulations using STAR-CD were also perform ed to evaluate effects of intake duct geometry and structure variation inside air cleaner on the negative pressure distribution of overall intake system. Studies for improving the back pressure distribution have been proposed and quantitatively examined based on intensive case studies.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.15 no.4
• /
• pp.186-192
• /
• 2007

We research into the exhaust system volume what proving the optimum exhaust performance and combustion characteristics. Many automobile manufactures have developed complex exhaust system for environment regulation and noise reduction. This complex exhaust system provides acoustics silencing and low frequency noise for customers demand. Recently, automobile exhaust system have made the Dual muffler concerning to the noise and vibration reduction. Also it bring the engine performance down by decreasing the back pressure and temperature in the exhaust system. The experiments are carried out different volume of exhaust system. In order to establish the optimized conditions design factors which are taking exhaust system volume, it show how the exhaust performance influence on the engine performance in idling.

• Journal of Information Technology Applications and Management
• /
• v.24 no.1
• /
• pp.157-167
• /
• 2017

This paper proposes a simple sound source localization (SSL) method based on signal energies comparison and partial cross correlation for TDOA computation. Many sound source localization methods include multiple TDOA computations in order to eliminate front-back confusion. Multiple TDOA computations however increase the methods' computation times which need to be as minimal as possible for real-time applications. Our aim in this paper is to achieve the same results of localization using fewer computations. Using three microphones, we first compare signal energies to predict which quadrant the sound source is in, and then we use partial cross correlation to estimate the TDOA value before computing the azimuth value. Also, we apply a threshold value to reinforce our prediction method. Our experimental results show that the proposed method has less computation time; spending approximately 30% less time than previous three microphone methods.

• Transactions of the Society of Information Storage Systems
• /
• v.6 no.1
• /
• pp.18-24
• /
• 2010

The performance of the speech recognition mainly depends on uncertain factors such as speaker's conditions and environmental effects. The present study deals with the speech recognition of a number of multi-syllable isolated Korean words using soft computing techniques such as back-propagation neural network, fuzzy inference system, and fuzzy neural network. Feature patterns for the speech recognition are analyzed with 12th order thirty frames that are normalized by the linear predictive coding and Cepstrums. Using four models of speech recognizer, actual experiments for both single-speakers and multiple-speakers are conducted. Through this study, the recognizers of combined fuzzy logic and back-propagation neural network and fuzzy neural network show the better performance in identifying the speech recognition.

• Korean Journal of Optics and Photonics
• /
• v.19 no.1
• /
• pp.43-47
• /
• 2008

Nowadays, the pattern size of BLU (Back Light Unit) adopted in TFT-LCD (Thin Film Transistor Liquid Crystal Display) is typically a few tens of micrometers. However, recently, researches on the TFT-LCD output characteristics depending on various types of BLU patterns are being performed in order to improve the output and uniformity. In this study, we analyzed the influence of pattern size, distribution, and areal ratio on the output characteristics.

• Journal of Digital Contents Society
• /
• v.5 no.3
• /
• pp.245-249
• /
• 2004

This paper shows the reconstruction of existing composition chord program using back propagation neural network. In this approach, in order to produce the expectation values, weight values are controlled by neural network which rued chord pattern as a input vector. Experimental results showed that proposed approach is superior to a popular chord pattern method rather than existing composition program.

• Journal of Institute of Control, Robotics and Systems
• /
• v.7 no.3
• /
• pp.218-228
• /
• 2001

The switched reluctance motor(SRM) has been increasingly used in high-performance servo applications such as electric vehicles, aircraft, and direct-drive robots. The dynamic equations of SRMs are, however, highly nonlinear and this makes it difficult to control SRMs with high performance. In this paper, we propose a new robust current tracking controller for SAMs which can compensate the nonlinear characteristics of SRM(i.e., back-emf and inductance) completely and hence shows perfect tracking performance even with an arbitrary small current control loop gain. Furthermore, even in case that there exist some model uncertainties, our current controller guarantees that the stator currents can track the reference current commands with sufficiently small tracking errors. In order to justify our work, we present the tracking performance analysis and some simulation results.