• Structural Engineering and Mechanics
• /
• v.26 no.6
• /
• pp.709-723
• /
• 2007

Considering the vast usage of time-history dynamic analyses to calculate structural responses and lack of sufficient and suitable earthquake records, generation of artificial accelerograms is very necessary. The main target of this paper is to present a novel method based on nonstationary Kanai-Tajimi model and wavelet transform to generate more artificial earthquake records, which are compatible with target spectrum. In this regard, the generalized nonstationary Kanai-Tajimi model to include the nonstationary evaluation of amplitude and dominant frequency of ground motion and properties of wavelet transform is used to generate ground acceleration time history. Application of the method for El Centro 1940 earthquake and two Iranian earthquakes (Tabas 1978 and Manjil 1990) is presented. It is shown that the model and identification algorithms are able to accurately capture the nonstationary features of these earthquake accelerograms. The statistical characteristics of the spectral response of the generated accelerograms are compared with those for the actual records to demonstrate the effectiveness of the method. Also, for comparison of the presented method with other methods, the response spectra of the synthetic accelerograms compared with the models of Fan and Ahmadi (1990) and Rofooei et al. (2001) and it is shown that the response spectra of the synthetic accelerograms with the method of this paper are close to those of actual earthquakes.

• Journal of Korea Multimedia Society
• /
• v.20 no.8
• /
• pp.1299-1311
• /
• 2017

Sounds based machine fault diagnosis recovers all the studies that aim to detect automatically faults or damages on machines using the sounds emitted by these machines. Conventional methods that use mathematical models have been found inaccurate because of the complexity of the industry machinery systems and the obvious existence of nonlinear factors such as noises. Therefore, any fault diagnosis issue can be treated as a pattern recognition problem. We present here an automatic fault diagnosis system of hand drills using discrete wavelet transform (DWT) and pattern recognition techniques such as principal component analysis (PCA) and artificial neural networks (ANN). The diagnosis system consists of three steps. Because of the presence of many noisy patterns in our signals, we first conduct a filtering analysis based on DWT. Second, the wavelet coefficients of the filtered signals are extracted as our features for the pattern recognition part. Third, PCA is performed over the wavelet coefficients in order to reduce the dimensionality of the feature vectors. Finally, the very first principal components are used as the inputs of an ANN based classifier to detect the wear on the drills. The results show that the proposed DWT-PCA-ANN method can be used for the sounds based automated diagnosis system.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2006.05a
• /
• pp.1450-1454
• /
• 2006

Empirical mode decomposition (EMD) is applied to analyze time series characterized with nonlinearity and nonstationarity. This decomposition could be utilized to construct finite and small number intrinsic mode functions (IMF) that describe complicated time series, while admitting the Hilbert transformation properties. EMD has the capability of being adaptive, capture local characteristics, and applicable to nonlinear and nonstationary processes. Unlike discrete wavelet transform (DWT), IMF eliminates spurious harmonics and retains meaningful instantaneous frequencies. Examples based on data representing natural phenomena are given to demonstrate highlight the power of this method in contrast and comparison of other ones. A presentation of the energy-frequency-time distribution of these signals found to be more informative and intuitive when based on Hilbert transformation.

• Smart Structures and Systems
• /
• v.22 no.2
• /
• pp.175-183
• /
• 2018

Time-series data often contain one of the most valuable pieces of information in many fields including manufacturing. Because time-series data are relatively cheap to acquire, they (e.g., vibration signals) have become a crucial part of big data even in manufacturing shop floors. Recently, deep-learning models have shown state-of-art performance for analyzing big data because of their sophisticated structures and considerable computational power. Traditional models for a machinery-monitoring system have highly relied on features selected by human experts. In addition, the representational power of such models fails as the data distribution becomes complicated. On the other hand, deep-learning models automatically select highly abstracted features during the optimization process, and their representational power is better than that of traditional neural network models. However, the applicability of deep-learning models to the field of prognostics and health management (PHM) has not been well investigated yet. This study integrates the "residual fitting" mechanism inherently embedded in the wavelet transform into the convolutional neural network deep-learning structure. As a result, the architecture combines a signal smoother and classification procedures into a single model. Validation results from rotor vibration data demonstrate that our model outperforms all other off-the-shelf feature-based models.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.10
• /
• pp.2121-2128
• /
• 2012

This paper proposed the real time face area detection using Wavelet transform and the strong detection algorithm that satisfies the efficiency of computation and detection performance at the same time was proposed. The detected face image recognizes the face by configuring the low-dimensional face symbol through the principal component analysis. The proposed method is well suited for real-time system construction because it doesn't require a lot of computation compared to the existing geometric feature-based method or appearance-based method and it can maintain high recognition rate using the minimum amount of information. In addition, in order to reduce the wrong recognition or recognition error occurred during face recognition, the input symbol of Hidden Markov Model is used by configuring the feature values projected to the unique space as a certain symbol through clustering algorithm. By doing so, any input face will be recognized as a face model that has the highest probability. As a result of experiment, when comparing the existing method Euclidean and Mahananobis, the proposed method showed superior recognition performance in incorrect matching or matching error.

• Structural Engineering and Mechanics
• /
• v.87 no.1
• /
• pp.19-28
• /
• 2023

This paper describes the experimental investigation of steel pallet rack beam-to-column connec-tions. Total behavior of moment-rotation (M-φ) curve and the effect of particular characteristics on the behavior of connection were studied and the associated load strain relationship and corre-sponding failure modes are presented. In this respect, an estimation of SPRBCCs moment and rotation are highly recommended in early stages of design and construction. In this study, a new approach based on Support Vector Machines (SVMs) coupled with discrete wavelet transform (DWT) is designed and adapted to estimate SPRBCCs moment and rotation according to four input parameters (column thickness, depth of connector and load, beam depth,). Results of SVM-WAVELET model was compared with genetic programming (GP) and artificial neural networks (ANNs) models. Following the results, SVM-WAVELET algorithm is helpful in order to enhance the accuracy compared to GP and ANN. It was conclusively observed that application of SVM-WAVELET is especially promising as an alternative approach to estimate the SPRBCCs moment and rotation.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.7A
• /
• pp.1037-1043
• /
• 2000

In this paper we present an object-based watermarking scheme, which cuts out several featured portions from an image and casts chaotic sequences as a watermark into the extracted object image in the discrete wavelet domain with respect to the models of the human visual system. In such a way we can insert watermark to several objects of an image separately. Advantages of the proposed scheme include that it can protect featured object out of an image selectively as well as entire image and casts watermark sequences according to human visual masking utilizing the time-frequency localization property of the wavelet transform.

• The KSFM Journal of Fluid Machinery
• /
• v.5 no.4 s.17
• /
• pp.47-53
• /
• 2002

The steam generator feedwater flow-rate in a nuclear power plant was estimated by means of artificial neural networks with the wavelet analysis for enhanced information extraction. The fouling of venturi meters, used for steam generator feedwater flow-rate in pressurized water reactors, may result in unnecessary plant power derating. The back-propagation network was used to generate models of signals for a pressurized water reactor Multiple-input, single-output hetero-associative networks were used for evaluating the feedwater flow rate as a function of a set of related variables. The wavelet was used as a low pass filter eliminating the noise from the raw signals. The results have shown that possible fouling of venturi can be detected by neural networks, and the feedwater flow-rate can be predicted as an alternative to existing methods. The research has also indicated that the decomposition of signals by wavelet transform is a powerful approach to signal analysis for denoising.

• Journal of Korea Water Resources Association
• /
• v.31 no.3
• /
• pp.317-326
• /
• 1998

Water consumption characteristics on the northern part of Seoul were analyzed using wavelet transform with a base function of Coiflets 5. It turns out that long term evolution mode detected at 212 scale in 1995 was in a shape of hyperbolic tangent over the entire period due to the development of Sanggae resident site. Furthermore, there was seasonal water demand having something to do with economic cycle which reached its peak at the ends of June and December. The amount of this additional consumption was about $1,700\;\textrm{cm}^3/hr$ on June and $500\;\textrm{cm}^3/hr$ on December. It was also shown that the periods of energy containing sinusoidal component were 3.13 day, 33.33 hr, 23.98 hr and 12 hr, respectively, and the amplitude of 23.98 hr component was the most humongous. The components of relatively short frequency detected at $2^i$[i = 1,2,…12] scale were following Gaussian PDF. The most reliable predictive models are multiple AR[32,16,23] and ARMA[20, 16, 10, 23] which the input of temperature from the view point of minimized predictive error, mutual independence or residuals and the availableness of reliable meteorological data. The predicted values of water supply were quite consistent with the measured data which cast a possibility of the deployment of the predictive model developed in this study for the optimal management of water supply facilities.

• Journal of Korea Water Resources Association
• /
• v.41 no.10
• /
• pp.1035-1044
• /
• 2008

It is critical to study on data charateristics analysis and prediction for the flood disaster prevention and water quality monitoring because discharge and TOC data in a river channel are strongly nonlinear. Therefore, in the present study, prediction models for discharge, TOC, and TOC load data were developed using approximation component in the last level and detail components segregated by wavelet transform. The results show that the developed model overcame the persistence phenomenon which could be seen from previous models and improved the prediciton accuracy comparing with the previous models. It might be expected that the results from the present study can mitigate flood disaster damage and construct active alternatives to various water quality problems in the future.