• 음성과학
• /
• 제11권1호
• /
• pp.7-20
• /
• 2004

In this paper, we propose a two-step noise compensation algorithm in feature extraction for achieving robust speech recognition. The proposed method frees us from requiring a priori information on noisy environments and is simple to implement. First, in frequency domain, the Harmonics-based Spectral Subtraction (HSS) is applied so that it reduces the additive background noise and makes the shape of harmonics in speech spectrum more pronounced. We then apply a judiciously weighted variance Feature Vector Normalization (FVN) to compensate for both the channel distortion and additive noise. The weighted variance FVN compensates for the variance mismatch in both the speech and the non-speech regions respectively. Representative performance evaluation using Aurora 2 database shows that the proposed method yields 27.18% relative improvement in accuracy under a multi-noise training task and 57.94% relative improvement under a clean training task.

• 한국지반공학회:학술대회논문집
• /
• 한국지반공학회 1999년도 연약지반처리위원회 학술세미나
• /
• pp.14-22
• /
• 1999

Daylight collapse have been occurred by about 6.$^{0}$ m deep at ground surface which connected to the ground surface and excessive overbreak have been occurred by the space and height of 3.$^1$~6.$^2$m at crown head part of the tunnel during tunnelling of lower-half part after completing upper-half part on tunnelling of a phyllite mountain by NATM method at the construction work of two way-double track national road. This study is a successful illustration case of earth improvement by confirming structural safety of the tunnel in a whole through solving the cause of the tunnel collapse and the work have completed successfully through applying such earth strengthening method as cement mortarㆍcement milk injection, S.G.R, steel pipe reinforced multi-step grouting etc.

• Journal of Mechanical Science and Technology
• /
• 제14권9호
• /
• pp.1005-1012
• /
• 2000

Numerical optimization techniques combined with a three-dimensional thin-layer Navier-Stokes solver are presented to find an optimum shape of a stator blade in an axial compressor through calculations of single stage rotor-stator flow. Governing differential equations are discretized using an explicit finite difference method and solved by a multi-stage Runge-Kutta scheme. Baldwin-Lomax model is chosen to describe turbulence. A spatially-varying time-step and an implicit residual smoothing are used to accelerate convergence. A steady mixing approach is used to pass information between stator and rotor blades. For numerical optimization, searching direction is found by the steepest decent and conjugate direction methods, and the golden section method is used to determine optimum moving distance along the searching direction. The object of present optimization is to maximize efficiency. An optimum stacking line is found to design a custom-tailored 3-dimensional blade for maximum efficiency with the other parameters fixed.

• Water Engineering Research
• /
• 제4권2호
• /
• pp.69-85
• /
• 2003

Previously developed empirical equations used to calculate the parameters of the transient storage model are analyzed in depth in order to evaluate their behavior in representing solute transport in the natural streams with storage zone. A comparative analysis of the existing theoretical and experimental equations used to predict parameters of the transient storage (TS) model is reported. New simplified equations for predicting 4 key parameters of the TS model using hydraulic data sets that are easily obtained in the natural streams are also developed. The weighted one-step Huber method, which is one of the nonlinear multi-regression methods, is applied to derive new parameters equation. These equations are proven to be superior in explaining mixing characteristics of natural streams with the transient storage zone more precisely than the other existing equations.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2004년도 추계학술대회 논문집
• /
• pp.1113-1118
• /
• 2004

This paper presents the dynamic analysis method for an electromechanical system. The engineer has at his disposal a variety of software simulation tools. However, difficulties arise when the study of the behavior of complex electromechanical systems in combination with coupling element is required. Typical examples of such systems are machines for factory automation, home automation, and office automation. Dynamic systems analysis packages or electronic systems analysis packages offer the restrictive to simulate these mixed systems such electromechanical product. Electronic circuit analysis algorithm is easily incorporated into a multi-body dynamics analysis algorithm. The governing equation of electronic circuit is formulated as a differential algebraic equation form including both electrical and mechanical variables and is simultaneously solved in every time step. This analysis method clearly demonstrates the application potential for mixed electromechanical simulation.

• Structural Engineering and Mechanics
• /
• 제28권5호
• /
• pp.549-570
• /
• 2008

Numerical integration is an efficient approach for nonlinear dynamic analysis. In this paper, general category of the implicit integration errors will be discussed. In order to decrease the errors, Dynamic Relaxation method with modified time step (MFT) will be used. This procedure leads to an alternative algorithm which is very general and can be utilized with any implicit integration scheme. For numerical verification of the proposed technique, some single and multi degrees of freedom nonlinear dynamic systems will be analyzed. Moreover, results are compared with both exact and other available solutions. Suitable accuracy, high efficiency, simplicity, vector operations and automatic procedures are the main merits of the new algorithm in solving nonlinear dynamic problems.

• 한국CDE학회논문집
• /
• 제4권2호
• /
• pp.100-109
• /
• 1999

The bearing design includes the steps of selection bering type, selection bearing subtype, and determining the peripheral equipments. In this paper decision making methodologies are compared to propose a stepwise decision methodology to the bearing selection problem. An artificial neural network trained with design cases is used for selecting a bearing type in the first step. Then the subtype of the bearing is selected using the weighting method, high is a kind of multi-criteria decision making method. Finally, the types of peripheral equipments such as lubrication devices, seals and bearing housings are determined using a rule-based expert system.

• Journal of electromagnetic engineering and science
• /
• 제2권1호
• /
• pp.11-15
• /
• 2002

In this paper, we have proposed the new configuration of wideband antenna using CPW fred lines. The proposed antenna has CPW feed lines and U-slot to achieve wade bandwidth with good impedance matching. The use of CPW feed line can decrease the number of substrates. It is compared with the conventional antenna find by a microstrip fled line. The parameters of CPW fred lines were studied by using the quasi-static approximation which is based on the conformal mapping method. The analysis of CPW discontinuities such as the tapered-step structure and the open ended gap were studied by using the quasi-static approximation which is based on the boundary element method. Also, the equivalent circuit model of multi-layer antenna were proposed using the cavity model. Experiment results of the proposed antenna 7how wide-bandwidth characteristics and good radiation patterns in passband.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1989년도 한국자동제어학술회의논문집; Seoul, Korea; 27-28 Oct. 1989
• /
• pp.635-642
• /
• 1989

We formulated the multi-robot trajectory problem into a series of NLP problem, each of which is that of finding the optimal tip positions of the robots for the next time step. The NLP problem is composed of an objective function and three constraints, namely: a) Joint position limits, b) Joint velocity limits, and c) Collision-avoidance constraints. By solving a series of NLP problem, optimally coordinated trajectories can be determined without requiring any prior path information. This is a novel departure from the previous approach in which either all paths or at least one path is assumed to be given. Practical application of the developed method is for optimal synthesis of multiple robot trajectories in off-line. To test the validity and effectiveness of the method, numerical examples are illustrated.

• 한국정밀공학회지
• /
• 제26권12호
• /
• pp.62-68
• /
• 2009

In this paper, we investigate a new method to measure the thickness of thin transparent objects utilizing a step index multi-mode optical fiber sensor. The method mainly depends on the refraction rate of transparent target, the diameter of optical fibers and the distance to reflector. We confirmed the effects of these parameters through the experimental verification tests. The comparison between the theoretical vs. analytical results shows good agreements with each other. The proposed model also enables users to measure the thickness of thin transparent objects without considering the reflection from the target. This approach provides simple, cost-effective and non-contact solutions to measure the thickness.