• The Journal of the Acoustical Society of Korea
• /
• 제29권1호
• /
• pp.41-47
• /
• 2010

The pitch alteration technique for voice conversion is classified in time domain, frequency domain and hybrid domain. The Hybrid domain method has a merit of clearness and natural-ness of pitch altered speech but has the major drawback of long processing time. In this paper, we proposed a new method that can reduce the processing time of pitch alteration in time-frequency hybrid domain. We omitted the bit-reversing process of FFT and IFFT in changing the processing domain. Therefore we can reduce the processing time by 86.26% to the conventional method with same quality.

• Journal of Intelligence and Information Systems
• /
• 제5권1호
• /
• pp.103-123
• /
• 1999

Input filtering as a preprocessing method is so much crucial to get good performance in time series forecasting. There are a few preprocessing methods (i.e. ARMA outputs as time domain filters, and Fourier transform or wavelet transform as time-frequency domain filters) for handling time series. Specially, the time-frequency domain filters describe the fractal structure of financial markets better than the time domain filters due to theoretically additional frequency information. Therefore, we, first of all, try to describe and analyze specially some issues on the effectiveness of different filtering methods from viewpoint of the performance of a neural network based forecasting. And then we discuss about neural network model architecture issues, for example, what type of neural network learning architecture is selected for our time series forecasting, and what input size should be applied to a model. In this study an input selection problem is limited to a size selection of the lagged input variables. To solve this problem, we simulate on analyzing and comparing a few neural networks having different model architecture and also use an embedding dimension measure as chaotic time series analysis or nonlinear dynamic analysis to reduce the dimensionality (i.e. the size of time delayed input variables) of the models. Throughout our study, experiments for integration methods of joint time-frequency analysis and neural network techniques are applied to a case study of daily Korean won / U. S dollar exchange returns and finally we suggest an integration framework for future research from our experimental results.

• Journal of Industrial Technology
• /
• 제22권B호
• /
• pp.163-167
• /
• 2002

Autonomic nervous system play an important role of keeping our health as balancing homeostasis. But the abnormality of these abilities makes our presence be feeble. To obtain these information of body which helps for us to decide whether one is healthy or not, based on the study of Heart Rate Variability. In this paper, we presented HRV model and its processing steps to extract some information of human body. After that, some experimental results are presented in time-domain and frequency-domain.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• 제16권12호
• /
• pp.1224-1230
• /
• 2006

Pipe thinning is one of the major issues for the structural fracture of pipes of nuclear power plants. Therefore a method to inspect a large area of piping systems quickly and accurately is needed. In this paper, we proposed the method for monitoring pipe thinning. Our basic idea come from that a group velocity of impact wave is different as wall thickness. If the group velocity is measured, wall thickness can be estimated. To obtain the group velocity, time -frequency analysis is used. This is because an arrival time difference can be measured easily in time-frequency domain rather than time domain. To test the performance of this technique, experiments have been performed for a plate and U type pipe. Results show that the proposed technique is quite powerful in the monitoring pipe thinning.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 한국지진공학회 2001년도 추계 학술발표회 논문집 Proceedings of EESK Conference-Fall 2001
• /
• pp.36-43
• /
• 2001

In Korea, we are absolutely short of earthquake data in good quality from moderate and large earthquakes, which are needed fur the study of strong ground motion characteristics. This means that the best use of the available data is needed far the time being. In this respect, several methods are suggested in this paper, which can be applied in the process of data selection and analysis. First, it is shown that the calibration status of seismic stations can be easily checked by comparing the spectra from accelerometer and velocity sensor both of which are located at the same location. Secondly, it is recommended that S/N ratio in the frequency domain should be checked before discarding the data by only look of the data in time domain. Thirdly, the saturated earthquake data caused by ground motion level exceeding the detection limit of a seismograph are considered to see if such data can be used for spectrum analysis by performing numerical simulation. The result reveals that the saturated data can still be used within the dominant frequency range according to the levels of saturation. Finally, a technique to minimize the window effect that distorts the low frequency spectrum is suggested. This technique involves detrending in displacement domain once the displacement data are obtained by integration of low frequency components of the original data in time domain. Especially, the low frequency component can be separated by using discrete wavelet transform among many alternatives. All of these methods mentioned above may increase the available earthquake data and frequency range.

• Journal of Ship and Ocean Technology
• /
• 제7권3호
• /
• pp.34-55
• /
• 2003

Explicit Kutta condition approximation, proved useful in existing time-domain solver of the unsteady propeller problem, requires a specified functional behavior of the vorticity in space near the trailing edge. In this paper, the strength of the discrete vortices is controlled to have a specified behavior in space in the frequency domain approach. A new formulation is introduced and is implemented for analysis of a lifting surface of a rectangular planform. Sample computations carried out according to the new formulation compares well with that of existing unsteady lifting problem in the time domain.

• Proceedings of the Korean Nuclear Society Conference
• /
• 한국원자력학회 1996년도 춘계학술발표회논문집(1)
• /
• pp.587-593
• /
• 1996

Vibration analysis is one of the most powerful tools available for the detection and isolation of incipient faults in mechanical systems. The methods of vibration analysis in use today and under continuous study are broad band vibration monitoring, time domain analysis, and frequency domain analysis. In recent years, great interest has been generated concerning the use of time-frequency representation and its application for a machinery diagnostics and condition monitoring system. The objective of the research described in this paper was to develop a new diagnostic tool for the rotating machinery. This paper introduces a new time-frequency representation, Directional Wigner-Ville Distribution, which analyses the time- frequency structure of the rotating machinery vibration.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• 제37권4호
• /
• pp.313-320
• /
• 2001

In the presence of incident waves with different frequencies, there are second order sum and difference frequency wane exciting forces due to the nonlinearty of the incident waves. Although the magnitudes of these nonlinear forces are small, they act at sum and difference frequencies away from those of the incident waves. So, the second order sum and difference frequency wane loads occurring close to the natural frequencies of TLPs often give greater contributions to high and low frequency resonant responses. The components of the second order forces which depend on first order quantities have been evaluated using the three dimensional source distribution method. The numerical results of time domain motion analysis for the nonlinear wave exciting forces in regular waves are compared with the numerical ones of frequency domain analysis. The results of comparison confirmed the validity of the proposed approach.

• Earthquakes and Structures
• /
• 제6권3호
• /
• pp.251-266
• /
• 2014

Dynamic analyses for a suite of ground of motions were conducted on concrete gravity dam sections to examine the earthquake induced stresses and effective damping. For this purpose, frequency domain methods that rigorously incorporate dam-reservoir-foundation interaction and time domain methods with approximate hydrodynamic foundation interaction effects were employed. The maximum principal tensile stresses and their distribution at the dam base, which are important parameters for concrete dam design, were obtained using the frequency domain approach. Prediction equations were proposed for these stresses and their distribution at the dam base. Comparisons of the stress results obtained using frequency and time domain methods revealed that the dam height and ratio of modulus of elasticity of foundation rock to concrete are significant parameters that may influence earthquake induced stresses. A new effective damping prediction equation was proposed in order to estimate earthquake stresses accurately with the approximate time domain approach.

• Proceedings of the Korean Society of Marine Engineers Conference
• /
• 한국마린엔지니어링학회 2002년도 춘계학술대회논문집
• /
• pp.213-217
• /
• 2002

Recently the subject of "wavelet analysis" has be drawn by both mathematical and engineering application fields such as Signal Processing, Compression/Decomposition, Wavelet-Neural Network, Statistics and etc. Even though its similar to courier analysis, wavelet is a versatile tool with much mathematical content and great potential for applications. Especially, wavelet transform uses localizable various mother wavelet functions in time-frequency domain. Therefore, wavelet transform has good time-analysis ability for high frequency component, and has good frequency-analysis ability for low frequency component. Using the discriminative ability is more easy method than other conventional techniques. In this paper, Morlet wavelet transform was applied to discriminate the kind of line fault by acquired data from real power transformation network. The experimental result presented that Morlet wavelet transform is easier, and more useful method than the FFW (Fast courier Transform).ransform).