• 인터넷정보학회논문지
• /
• 제10권5호
• /
• pp.127-133
• /
• 2009

본 논문은 가중 퍼지소속함수 기반 신경망(Neural Network with Weighted Fuzzy Membership Functions, NEWFM)을 이용하여 간질 증세를 가진 사람과 건강한 사람의 뇌파(electroencephalogram, EEG)로부터 정상 파형과 간질(epilepsy) 파형을 분류하는 방안을 제시하고 있다. NEWFM에서 사용할 특징입력을 추출하기 위해서 첫 번째 단계에서는 웨이블릿 변환(wavelet transform, WT)을 이용하였다. 두 번째 단계에서는 첫 번째 단계에서 생성한 웨이블릿 계수들을 주파수 분포와 주파수 변동량을 이용하여 24개의 특징입력을 추출하였다. NEWFM은 이들 24개의 특징입력을 이용하여 정상 파형과 간질 파형을 분류하였을 때 98%의 분류성능을 나타내었다.

• 전기학회논문지P
• /
• 제57권2호
• /
• pp.91-95
• /
• 2008

The large AC generator fault may lead to large impacts or perturbations in power system. The generator protection control systems in Korea have been imported and operated through a turn-key from overseas entirely. Therefore a study of the generator protection field has in urgent need for a stable operation of the imported goods. In present, the algorithm using the current ratio differential relaying based DFT for stator winding protection or a fault detection had been applied that of internal fault protection of a generator. the DFT used for the analysis of transient state signal conventionally had defects losing a time information in the course of transforming a target signal to frequency domain. In this paper, the discrete wavelet transform (DWT) was applied a fault detection of the generator being superior to a transient state signal analysis and being easy to real time realization. The fault signals after executing a terminal fault modeling collect using a MATLAB package, and calculate the wavelet coefficients through the process of a muiti-level decomposition (MLD). The proposed algorithm for a fault detection using the Daubechies WT (wavelet transform) was executed with a C language and the commend line function for the real time realization after analyzing MATLAB's graphical interface. The advanced technique had improved faster a speed of fault discrimination than a conventional DFR based on DFT.

• 인터넷정보학회논문지
• /
• 제10권2호
• /
• pp.179-186
• /
• 2009

본 논문은 가중 퍼지소속함수 기반 신경망(Neural Network with Weighted Fuzzy Membership Functions, NEWFM)을 이용하여 Creighton University Ventricular Tachyarrhythmia DataBase(CUDB)의 심전도(ECG) 신호로부터 정상리듬(Normal Sinus Rhythm, NSR)과 심실빈맥/심실세동(Ventricular Tachycardia/Ventricular Fibrillation, VT/VF)을 분류하는 방안을 제시하고 있다. NEWFM에서 사용할 특징입력을 추출하기 위해서 첫 번째 단계에서는 웨이블릿 변환(wavelet transform, WT)을 이용하였다. 두 번째 단계에서는 첫 번째 단계에서 생성된 웨이블릿 계수들을 위상공간 재구성(Phase Space Reconstruction, PSR)과 첨단(Peak) 추출 기법의 입력 값으로 이용하여 2개의 특징입력을 추출하였다. NEWFM은 이들 2개의 특징입력을 이용하여 정상리듬과 심실빈맥/심실세동을 분류하였고 그 결과로 90.13%의 분류성능을 나타내었다.

• 한국컴퓨터정보학회논문지
• /
• 제29권5호
• /
• pp.21-29
• /
• 2024

CT 촬영 시 방사선량을 줄이면 피폭 위험성을 낮출 수 있으나, 영상 해상도가 크게 저하 될 뿐아니라 잡음(noise) 발생으로 인해 진단의 효용성이 떨어진다. 따라서, CT 영상에서의 잡음제거는 영상복원 분야에 있어 매우 중요하고 필수적인 처리 과정이다. 영상 영역에서 잡음과 원래 신호를 분리하여 잡음만을 제거하는 것은 한계가 있다. 본 논문에서는 웨이블릿 변환 기반 GAN 모델 즉, WT-GAN(wavelet transform-based GAN) 모델을 이용하여 CT 영상에서 효과적으로 잡음 제거하고자 한다. 여기서 사용된 GAN 모델은 U-Net 구조의 생성자와 PatchGAN 구조의 판별자를 통해 잡음제거 영상을 생성한다. 본 논문에서 제안된 WT-GAN 모델의 성능 평가를 위해 다양한 잡음, 즉, 가우시안 잡음(Gaussian noise), 포아송 잡음 (Poisson noise) 그리고 스펙클 잡음 (speckle noise)에 의해 훼손된 CT 영상을 대상으로 실험하였다. 성능 실험 결과, WT-GAN 모델은 전통적인 필터 즉, BM3D 필터뿐만 아니라 기존의 딥러닝 모델인 DnCNN, CDAE 모형 그리고 U-Net GAN 모형보다 정성적이고, 정량적인 척도 즉, PSNR (Peak Signal-to-Noise Ratio) 그리고 SSIM (Structural Similarity Index Measure) 면에서 우수한 결과를 보였다.

• Nuclear Engineering and Technology
• /
• 제35권5호
• /
• pp.454-470
• /
• 2003

In this paper, a preliminary study for development of a fault diagnosis is presented for monitoring and diagnosing faults in the agitator driving system of a high temperature reduction reactor. In order to identify a fault occurrence and classify the fault cause, vibration signals measured by accelerometers on the outer shroud of the agitator driving system are firstly decomposed by wavelet transform (WT) and the features corresponding to each fault type are extracted. For the diagnosis, the fuzzy ARTMAP is employed and thereby, based on the features extracted from the WT, the robust fault classifier can be implemented with a very short training time - a single training epoch and a single learning iteration is sufficient for training the fault classifier. The test results demonstrate satisfactory classification for the faults pre-categorized from considerations of possible occurrence during experiments on a small-scale reduction reactor.

• 한국해양공학회지
• /
• 제22권3호
• /
• pp.71-75
• /
• 2008

This paper illustrates haw $Y_2O_3$ contributes to crack-healing strengths as a function of crack-healing temperature and the additive amount. In investigating mechanical properties, the indentation fracture method is very simple and useful, but careful attention must be paid to the statistical data processing because data may be scattered excessively, especially for brittle materials. To estimate accurate AE signal properties we applied the useful time-frequency method with a discrete wavelet analysis algorithm. In experiments, three kinds of specimens were prepared. After the specimens were indented by a Vickers indentor, they were heat-treated and crack-healed to evaluate bending strength and the AE signal. With higher amounts of the additive powder, as 1, 3, or 5% wt. of $Y_2O_3$, the concentrative tendency of dominant frequency trended toward lower frequency groups. The $Al_2O_3$ ceramic with 3% wt. of $Y_2O_3$ was judged most suitable because it demonstrated superior crack-healing ability and relative concentration on the highest frequency group.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권8호
• /
• pp.3862-3879
• /
• 2019

More and more problems for public security have occurred due to the limited solutions for drone detection especially for micro-drone in long range conditions. This paper aims at dealing with drones detection using a radar system. The radio frequency (RF) signals emitted by a controller can be acquired using the radar, which are usually too weak to extract. To detect the drone successfully, the static clutters and linear trend terms are suppressed based on the background estimation algorithm and linear trend suppression. The principal component analysis technique is used to classify the noises and effective RF signals. The automatic gain control technique is used to enhance the signal to noise ratios (SNR) of RF signals. Meanwhile, the empirical mode decomposition (EMD) based wavelet transform (WT) is developed to decrease the influences of the Gaussian white noises. Then, both the azimuth information between the drone and radar and the bandwidth of the RF signals are acquired based on the statistical analysis algorithm developed in this paper. Meanwhile, the proposed accumulation algorithm can also provide the bandwidth estimation, which can be used to make a decision accurately whether there are drones or not in the detection environments based on the probability theory. The detection performance is validated with several experiments conducted outdoors with strong interferences.

• 한국유체기계학회 논문집
• /
• 제15권6호
• /
• pp.31-38
• /
• 2012

Rotor of rotating machinery is the highly damaged part. Fault of 7 different types was confirmed as the main causes of rotor damage from the pump failure history data in domestic and U.S. nuclear. For each fault types, simulation testing was performed and fault signals were collected form the sensors. To calculate the statistical parameters of time-domain & frequency-domain, measured signals were analyzed by using the discrete wavelet transform, fast fourier transform, statistical analysis. Total 84 parameters were obtained. And Effectiveness factor were used to evaluate the discrimination capacity of each parameter. From the effectiveness factor, RAW-P4/RAW-P7/WT2-NNL/WT2-EE/WT1-P1 showed high ranking. Finally, these parameters were selected as the feature parameters of intelligent fault diagnostics for rotor.

• 대한의용생체공학회:학술대회논문집
• /
• 대한의용생체공학회 1996년도 춘계학술대회
• /
• pp.248-251
• /
• 1996

In this paper, we presented WT(wavelet transform) domain feedback canceller in hearing aids system. Most transform methods produce blocking effect and this effect degrades the performance of feedback canceller and overall hearing aids system. As a solution we proposed WT based feedback canceller. The performance of this new approach was compared with LOT (lapped orthogonal transform) based method in the frequency domain. As a result, WT based feedback canceller has not shown the blocking effect and improved convergence rate as compared with the LOT based feedback canceller.

• 한국산업정보학회논문지
• /
• 제22권1호
• /
• pp.41-51
• /
• 2017

This Paper Describes a Method for the Evaluation of Sleep Apnea, Namely, the Peak Signal-to-noise ratio (PSNR) of Wavelet Transformed Electroencephalography (EEG) Data. The Purpose of this Study was to Investigate EEG Properties with Regard to Differences between Sleep Spindles and K-complexes and to Characterize Obstructive Sleep Apnea According to Sleep Stage. We Examined Non-REM and REM Sleep in 20 Patients with OSA and Established a New Approach for Detecting Sleep Apnea Base on EEG Frequency Changes According to Sleep Stage During Sleep Apnea Events. For Frequency Bands Corresponding to A3 Decomposition with a Sampling Applied to the KC and the Sleep Spindle Signal. In this Paper, the KC and Sleep Spindle are Ccalculated using MSE and PSNR for 4 Types of Mother Wavelets. Wavelet Transform Coefficients Were Obtained Around Sleep Spindles in Order to Identify the Frequency Information that Changed During Obstructive Sleep Apnea. We also Investigated Whether Quantification Analysis of EEG During Sleep Apnea is Valuable for Analyzing Sleep Spindles and The K-complexes in Patients. First, Decomposition of the EEG Signal from Feature Data was Carried out using 4 Different Types of Wavelets, Namely, Daubechies 3, Symlet 4, Biorthogonal 2.8, and Coiflet 3. We Compared the PSNR Accuracy for Each Wavelet Function and Found that Mother Wavelets Daubechies 3 and Biorthogonal 2.8 Surpassed the other Wavelet Functions in Performance. We have Attempted to Improve the Computing Efficiency as it Selects the most Suitable Wavelet Function that can be used for Sleep Spindle, K-complex Signal Processing Efficiently and Accurate Decision with Lesser Computational Time.