• Journal of Intelligence and Information Systems
• /
• v.5 no.1
• /
• 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 Broadcast Engineering
• /
• v.15 no.5
• /
• pp.617-630
• /
• 2010

At 37th VCEG meeting in Jan. 2009, Toshiba proposed Quadtree-based Adaptive Loop Filter (QALF). The basic concept of QALF is to apply Wiener filter to decoded image after the conventional deblocking filter and to represent the filter on/off flag data for each basic filtering unit in a more efficient way of quadtree structure. QALF could enhance the compression performance of around more than 9%, but the structure of one filter for a decoded frame leaves room for further improvement in the sense that optimal filter for one region of a frame could quite different from the optimal filter for other parts of a picture. This paper proposes multiple adaptive loop filters for better utilization of local characteristics of decoded frame to optimize the region-based Wiener filters. Additional filters, proposed in this paper, cover separate spatial area of each decoded frame according to the performance of previously designed filter(s) to provide the flexibility of rate-distortion based selection of the number of filters.

• Journal of radiological science and technology
• /
• v.44 no.6
• /
• pp.607-613
• /
• 2021

Long-bone examination is mainly used for inspection of the lower extremities. Recently, a long length detector (FXRD-1751S, VIEWORKS, Korea) with three digital detectors attached has been developed. High energy X-rays are used because pelvic areas require high image quality. In this case, X-rays are transmitted a lot in thin areas such as an ankle, and it is not suitable for diagnosing an image. Therefore, this study use copper filters made with 3D printers to increase image quality in the Long-bone inspection. A copper filter was manufactured in consideration of the overall thickness of the lower part. The experiment was conducted in anterior-posterior (AP) and lateral (LAT) positions, depending on the presence or absence of the filter. 5x5 pixels of region of interest (ROI) were selected from the pelvis, knee, and ankle areas. X-rays were irradiated under the conditions of 70 kVp and 40 mAs for AP, 80 kVp, and 63 mAs for lat when without filters, 90 kVp and 80 mAs for AP, 90 kVp and 100 mAs for lat when with filters. signal to noise ratio(SNR) ratio and contrast to noise (CNR) values were measured 1106.38, 14.34 before applying the filter and 1189.32, 70.43 after the filter. For the knee area, 650.44, 97.61 before applying the filter, and 1013.17, 444.24 after applying the filter. For the ankle area, 206.65, 23.68 before applying the filter and 993.50, 136.11 after applying the filter. In the Long-bone examination, SNR and CNR were greatly measured when the filter was applied, confirmed the availability of using the copper additional filter.

• ETRI Journal
• /
• v.26 no.4
• /
• pp.292-296
• /
• 2004

In this paper we present a new current-mode electronically tunable universal filter using only plus-type current controlled conveyors (CCCII+s) and grounded capacitors. The proposed circuit can simultaneously realize lowpass, bandpass, and highpass filter functions - all at high impedance outputs. The realization of a notch response does not require additional active elements. The circuit enjoys an independent current control of parameters $\omega_0$ and $\omega_0/Q$. No element matching conditions are imposed. Both its active and passive sensitivities are low.

• Korean Journal of Digital Imaging in Medicine
• /
• v.12 no.1
• /
• pp.19-25
• /
• 2010

This study is mainly for the reduction of exposure dose by using a heavy elements filter(Gd) in the digital radiology. They contained heavy elements filter of Gd and X-ray beam hardening filters such as Al and Cu. According to the results of experimental evaluation, X-ray property was not changed with variety of kVp in the case of the Gd filter. The surface dose and absorption dose were increased in order of Cu and Al. The contrast of image showed the higher value in order of Cu, Al and Gd. While the use of Gd has increased the numerical value of the CR image, and grayscale has decreased noise value of the DR image.

• Journal of Power Electronics
• /
• v.12 no.2
• /
• pp.326-332
• /
• 2012

This paper proposes a simple active damping algorithm for small-scale wind power systems with an LCL filter. Compared to an L filter or an LC filter, an LCL filter can decrease the harmonics induced by low switching frequencies and produce a satisfactory grid-side current using a comparatively low inductance. Additional active damping of the filter resonance is necessary when an LCL filter is used. This paper introduces an active damping method using a Discrete Fourier Transform (DFT) filter to improve performance without additional sensors or complexity. Experimental results are shown to verify the validity of the proposed algorithm as an active damping method.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.41 no.3
• /
• pp.311-319
• /
• 1992

In this paper, a direct adaptive pole placement controller for an unknown linear time-invariant single-input single-output nonminimum phase plant is proposed. To design this direct adaptive pole placement controller, the auxiliary signals are introduced. Consequently, a linear equation error model is formulated for estimating both the controller parameters and the additional auxiliary parameters. To estimate the controller parameters and the additional auxiliary parameters, the exponentially weighted least-squares algorithm is implemented, and a method of selecting the characteristic polynomials of the sensitivity function filters is proposed. In this method, all the past measurement data are weighted exponentially. A series of simulations for a nonminimum phase plant is presented to illustrate some features of both the parameter estimation and the output response of this adaptive pole placement controller.

• Journal of Electrical Engineering and Technology
• /
• v.13 no.1
• /
• pp.280-286
• /
• 2018

A sensorless active damping scheme for LCL filters in grid-connected parallel inverters for battery energy storage systems is proposed. This damping method is superior to the conventional notch filter and virtual damping methods with respect to robustness against the variation of the resonance of the filter and unnecessary additional current sensors. The theoretical analysis of the proposed damping method is explained in detail, along with the characteristic comparison to the conventional active damping methods. The performance verification of the proposed sensorless active damping method shows that its performance is comparable to that of the conventional virtual damping method, even without additional current sensors. Finally, simulation and experimental results are provided to examine the overall characteristics of the proposed method.

• Journal of radiological science and technology
• /
• v.17 no.2
• /
• pp.29-35
• /
• 1994

Tomographic examination requires several times of exposure usually, therefore, reduction of radiation dose per exposure without loss of image quality proves of great benefit to patient. We compared the exposure doses under the following experimental conditions, A and B. A is the combination of SRO 380(rare earth screen) and SRH(high contrast film) with additional Cu filters or without. B is the widely used combination which is BH-III and Cronex-4 without additional filter. As a result, comparing with the condition B, the condition A with additional filter of Cu 0.15 mm + Al 0.7 mm, Cu 0.4 mm+ Al 0.4 mm and Cu 0.8 mm showed better Image quality and lower surface dose, 44 %, 31 % and 24 % of the condition B, respectively.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.2
• /
• pp.367-376
• /
• 2016

This paper presents a cell balancing method for a single switch flyback converter with a multi-winding transformer. The conventional method using a flyback converter with a multi-winding transformer is simple and easy to control, but the voltage of each secondary winding coil might be non-uniform because of the unequal effective turn-ratio. In particular, it is difficult to control the non-uniform effect using turn-ratios because secondary coil has a limited number of turns. The non-uniform secondary voltages disturb the cell balancing procedure and induce an unbalance in cell voltages. Individual cell control by adding a switch for each cell can reduce the undesirable effect. However, the circuit becomes bulky, resulting in additional loss. The proposed method here uses the conventional flyback converter with an adjustment made to the output filters of the cells, instead of the additional switch. The magnitude of voltage applied to a particular cell can be reduced or increased according to the adjusted filter and the selected switching frequency. An analysis of the conventional converter configuration and the filter design method reveals the possibility of adequate cell balancing control without any additional switch on the secondary side.