• International Journal of Precision Engineering and Manufacturing
• /
• v.7 no.3
• /
• pp.14-17
• /
• 2006

To diagnosis abnormal compressor conditions in an air-conditioner, the acoustic emission (AE) signal, which is derived from wear condition, compressed air, and assembly error, was analyzed experimentally. Burst and continuous type AE signals resulted from metal contact and compressed air, and the raw AE signal of compressors was acquired in the production line. After extracting samples using waveforms, the Early Life Test (ELT) was conducted and the waveform was classified as normal or abnormal. Efficient parameters in the waveform pattern were investigated in time and frequency domains and a diagnosis algorithm for air-conditioners using Neural Network estimation is suggested.

• Journal of the Korean Society for Precision Engineering
• /
• v.21 no.11
• /
• pp.125-129
• /
• 2004

For the diagnosis of compressor abnormal condition in air-conditioner, AE signal which is derived from wear condition, compressed air and assembly error is analyzed experimentally. The burst and continuous type AE signal occurred by metal contact and compressed air and AE raw signal of compressors were directly acquired in production line. After extracting samples according to waveforms, Early Life Test(ELT) is conducted and classified to normal and abnormal waveform. The efficient parameters of waveform pattern are investigated in time and frequency domain and the diagnosis algorithm of air-conditioner by Neural Network estimation is suggested.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.1
• /
• pp.33-39
• /
• 2019

In this paper, we propose deep convolutional neural network(CNN) with bottleneck structure which improves the performance of earthquake classification. In order to address all possible forms of earthquakes including micro-earthquakes and artificial-earthquakes as well as large earthquakes, we need a representation and classifier that can effectively discriminate seismic waveforms in adverse conditions. In particular, to robustly classify seismic waveforms even in low snr, a deep CNN with 1x1 convolution bottleneck structure is proposed in raw seismic waveforms. The representative experimental results show that the proposed method is effective for noisy seismic waveforms and outperforms the previous state-of-the art methods on domestic earthquake database.

• MALSORI
• /
• no.56
• /
• pp.225-235
• /
• 2005

Speech unit concatenation with a large database is presently the most popular method for speech synthesis. In this approach, the mismatches at the unit boundaries are unavoidable and become one of the reasons for quality degradation. This paper proposes an algorithm to reduce undesired discontinuities between the subsequent units. Optimal matching points are calculated in two steps. Firstly, the fullback-Leibler distance measurement is utilized for the spectral matching, then the unit sliding and the overlap windowing are used for the waveform matching. The proposed algorithm is implemented for the corpus-based unit concatenating Korean text-to-speech system that has an automatically labeled database. Experimental results show that our algorithm is fairly better than the raw concatenation or the overlap smoothing method.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2011.04a
• /
• pp.572-577
• /
• 2011

The acoustic emission (AE) technique is a well-known non-destructive test technique, both in research and for industrial applications. It is mainly used to monitor the onset of cracking processes in materials and components. Predicting and preventing the crack phenomenon has attracted the attention of many researchers and has continued to provide a large incentive for the use of condition monitoring techniques to detect the earliest stages of cracks. In this research, goal is in grasping features of AE signal caused by crack growth. The envelope analysis with discrete wavelet transform (DWT) is used to find the characteristic of AE signal. To estimate feature of divided into three by crack length, the time waveform and the power spectrum were generated by the raw signals and the transferred signal processed by envelope analysis with DWT.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.441-446
• /
• 2020

In this paper, we propose the α-Feature Map Scaling (α-FMS) method which extends the FMS method that was designed to enhance the discriminative power of feature maps of deep neural networks in Speaker Verification (SV) systems. The FMS derives a scale vector from a feature map and then adds or multiplies them to the features, or sequentially apply both operations. However, the FMS method not only uses an identical scale vector for both addition and multiplication, but also has a limitation that it can only add a value between zero and one in case of addition. In this study, to overcome these limitations, we propose α-FMS to add a trainable parameter α to the feature map element-wise, and then multiply a scale vector. We compare the performance of the two methods: the one where α is a scalar, and the other where it is a vector. Both α-FMS methods are applied after each residual block of the deep neural network. The proposed system using the α-FMS methods are trained using the RawNet2 and tested using the VoxCeleb1 evaluation set. The result demonstrates an equal error rate of 2.47 % and 2.31 % for the two α-FMS methods respectively.

• Journal of Power System Engineering
• /
• v.7 no.2
• /
• pp.29-35
• /
• 2003

The monitoring and diagnostics of the rotating machinery have been received considerable attention for many years. The objectives are to classify the machinery condition and to find out the cause of abnormal condition. This paper describes a classification method of diagnosing the small reciprocating compressor for refrigerators using the artificial neural network and the wavelet transform. In order to extract salient features, the wavelet transform are used from primary noise signals. Since the wavelet transform decomposes raw time-waveform signals into two respective parts in the time space and frequency domain, more and better features can be obtained easier than time-waveform analysis. In the training phase for classification, self-organizing feature map(SOFM) and learning vector quantization(LVQ) are applied, and the accuracies of them ate compared with each other. This paper is focused on the development of an advanced signal classifier to automatize the vibration signal pattern recognition. This method is verified by small reciprocating compressors, for refrigerator and normal and abnormal conditions are classified with high flexibility and reliability.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2000.11a
• /
• pp.796-801
• /
• 2000

The monitoring and diagnostics of the rotating machinery have been received considerable attention for many years. The objectives are to classify the machinery condition and to find out the cause of abnormal condition. This paper describes a signal classification method for diagnosing the rotating machinery using the artificial neural network and the wavelet transform. In order to extract salient features, the wavelet transform are used from primary noise signals. Since the wavelet transform decomposes raw time-waveform signals into two respective parts in the time space and frequency domain, more and better features can be obtained easier than time-waveform analysis. In the training phase for classification, self-organizing feature map(SOFM) and learning vector quantization(LVQ) are applied, and the accuracies of them are compared with each other. This paper is focused on the development of an advanced signal classifier to automatise the vibration signal pattern recognition. This method is verified by small reciprocating compressors, for refrigerator and normal and abnormal conditions are classified with high flexibility and reliability.

• Journal of the korean academy of Pediatric Dentistry
• /
• v.43 no.2
• /
• pp.158-165
• /
• 2016

The purpose of this study was to analyze photoplethysmographic waveforms from pulse oximeter using raw data of red and infrared light and investigate the reference values of parameters (Height, Width50, Maximum slope, Minimum slope, Area) for evaluating pulpal blood flow in maxillary central incisors with normal pulp vitality in children. The study was performed in 30 pediatric patients, aged 7-16 years old, using pulse oximeter (MEKICS Co., Ltd, Korea) combined with a custom-made sensor. The raw data was obtained and recorded by custom-made software and analyzed by LabChart (v.7.3, ADInstruments, Germany) offline. In this study, we analyzed photoplethysmographic waveforms from pulse oximeter applied to maxillary central incisor for assessment of pulpal blood flow and suggested several reference values of young permanent maxillary central incisor with normal pulp. On average, the waveform of red light was higher, stiffer and wider than that of infrared light. Future studies about reference values for other normal teeth and the teeth with impaired pulp vitality are needed.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.5
• /
• pp.70-75
• /
• 2014

Super-resolution near-field structure (super-RENS) read-out samples are affected by a nonlinear and noncausal channel, which results in inter-symbol interference (ISI). In this study, we investigate asymmetry or domain bloom in super-RENS in terms of equalization. Domain bloom is caused by writing process in optical recording. We assume in this work that the asymmetry symbol conversion scheme is to generate asymmetric symbols, and then a linear finite impulse response filter can model the read-out channel. For equalizing this overall nonlinear channel, the read-out signals are deconvolved with the finite impulse response filter and its output is decided based on the decision rule table that is developed from the asymmetry symbol conversion scheme. The proposed equalizer is investigated with the simulations and the real super-RENS samples in terms of raw bit error rate.