• Proceedings of the KIEE Conference
• /
• 2003.10a
• /
• pp.236-238
• /
• 2003

The electron transport coefficients, the electron drift velocity W, the longitudinal diffusion coefficient $ND_L$ and $D_L/\mu$, in pure Ne were calculated over the wide E/N range from 0.01 to 800 Td at 1 Torr by two-term approximation of the Boltzmann equation for determination of electron collision cross sections set and for quantitative characteristic analysis of Ne molecular gas.

• International Journal of Ocean System Engineering
• /
• v.3 no.3
• /
• pp.158-163
• /
• 2013

The acoustic target strength (TS) is one of the most important parameters for a submarine's stealth design. Because modem submarines are larger than their predecessors, TS must be managed at each design stage in order to reduce it. To predict the TS of a submarine, TASTRAN R1 was developed based on a Kirchhoff approximation in a high-frequency range. This program can present TS values that include multi-bounce effect in the exterior and interior of the structure by combining geometric optics (GO) and physical optics (PO) methods, anechoic coating effect by using the reflection coefficient, and response time pattern for a detected target. In this paper, TS calculations for a submarine model with the above effects are simulated by using this developed program, and the TS results are discussed.

• The Journal of the Korea Contents Association
• /
• v.21 no.3
• /
• pp.616-625
• /
• 2021

Deep neural networks are an approximation method that approximates an arbitrary function to a linear model and then repeats additional approximation using a nonlinear active function. In this process, the method of evaluating the performance of approximation uses the loss function. Existing in-depth learning methods implement approximation that takes into account loss functions in the linear approximation process, but non-linear approximation phases that use active functions use non-linear transformation that is not related to reduction of loss functions of loss. This study proposes parametric activation functions that introduce scale parameters that can change the scale of activation functions and location parameters that can change the location of activation functions. By introducing parametric activation functions based on scale and location parameters, the performance of nonlinear approximation using activation functions can be improved. The scale and location parameters in each hidden layer can improve the performance of the deep neural network by determining parameters that minimize the loss function value through the learning process using the primary differential coefficient of the loss function for the parameters in the backpropagation. Through MNIST classification problems and XOR problems, parametric activation functions have been found to have superior performance over existing activation functions.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.26 no.11
• /
• pp.2390-2398
• /
• 2002

For integrity analysis of nuclear reactor pressure vessel, including the Pressurized thermal shock analysis, the fast and accurate calculation of the stress intensity factor at the crack tip is needed. For this, a simple approximation scheme is developed and the resulting stress intensity factors for axial semi-elliptical cracks in cylindrical vessel under various loading conditions are compared with those of the finite element method and other approximation methods, such as Raju-Newman's equation and ASME Sec. Xl approach. For these, three-dimensional finite-element analyses are performed to obtain the stress intensity factors for various surface cracks with t/R = 0.1. The approximation methods, incorporated in VINTIN (Vessel INTegrity analysis-INner flaws), utilizes the influence coefficients to calculate the stress intensity factor at the crack tip. This method has been compared with other solution methods including 3-D finite clement analysis for internal pressure, cooldown, and pressurized thermal shock loading conditions. The approximation solutions are within $\pm$2.5% of the those of FEA using symmetric model of one-forth of a vessel under pressure loading, and 1-3% higher under pressurized thermal shock condition. The analysis results confirm that the VINTIN method provides sufficiently accurate stress intensity factor values for axial semi-elliptical flaws on the surface of the reactor pressure vessel.

• Journal of the Korean Society of Radiology
• /
• v.8 no.1
• /
• pp.19-26
• /
• 2014

This paper suggests techniques to enhance coding time which is a problem in traditional fractal compression and to improve fidelity of reconstructed images by determining fractal coefficient through adaptive selection of block approximation formula. First, to reduce coding time, we construct a linear list of domain blocks of which characteristics is given by their luminance and variance and then we control block searching time according to the first permissible threshold value. Next, when employing three-level block partition, if a range block of minimum partition level cannot find a domain block which has a satisfying approximation error, we choose new approximation coefficients using a non-linear approximation of luminance term. This boosts the fidelity. Our experiment employing the above methods shows enhancement in the coding time more than two times over traditional coding methods and shows improvement in PSNR value by about 1-3dB at the same compression rate.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1829-1832
• /
• 2004

Closed loop stability of Extended Dynamic Matrix Control (EDMC) is investigated for limited sampling time. Linear approximation of the sensitivity functions is employed in the derivation of the stability condition. It is shown that the closed loop system will be stable if the control moves suppression coefficient ${\lambda}$ is taken arbitrarily large. Special cases such as M=P=1 and M=1, P>1 are discussed in more details.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2001.06a
• /
• pp.125-128
• /
• 2001

In this paper, We proposed the nonlinear controller and observer design for a ball and beam system. Unfortunately, for the ball and beam system, the control coefficient is zero whenever the angular velocity or ball position are zero. Therefore, the relative degree of the ball and beam system is not well defined. The presented the nonlinear controller and observer design is based on the approximation input-output feedback linearization. And we verified that the proposed nonlinear controller and observer scheme is the feasible through a computer simulation.

• Proceedings of the IEEK Conference
• /
• 2002.06d
• /
• pp.171-174
• /
• 2002

This paper presents an implementation method of phasor measurement device, which is based on the FPGA implementation of the sliding-DFT The design is verified by the timing simulation of its operation. The error effect of coefficient approximation and frequency deviation in the recursive implementation of the sliding-DFT is analytically derived and verified with the computer simulations.

• Journal of Astronomy and Space Sciences
• /
• v.19 no.4
• /
• pp.395-402
• /
• 2002

The NORAD type Two Line Element (TLE) was obtained from the osculating orbital elements by an iterative approximation procedure. The mathematical model was presented and computer program was developed for the conversion. The osculating orbital elements of the KOMPSAT-1 were converted into the NORAD TLE. Then the effect of the SGP4 atmospheric drag coefficient ($B^*$) was analyzed by comparison of the orbit propagation results with different $B^*$ values.

• Kyungpook Mathematical Journal
• /
• v.61 no.2
• /
• pp.383-393
• /
• 2021

In this paper, numerical techniques are presented for solving initial value problems of fractional differential equations with variable coefficients. The method is derived by applying a Taylor vector approximation. Moreover, the operational matrix of fractional integration of a Taylor vector is provided in order to transform the continuous equations into a system of algebraic equations. Furthermore, numerical examples demonstrate that this method is applicable and accurate.