Earticle

본 연구의 주요 목적은 경제시계열에 대한 다양한 시간주기에서의 분해를 실시할 수 있다는 소파동 분석의 장점을 활용하여, 주식수익률과 채권수익률을 분해하고 다양한 주기에 따른 경제변수의 의미를 살펴보는데 있다. 기존의 연구에서는 주식수익률과 채권수익률과의 관계가 정(+)의 관계인지, 부(-)의 관계인지에 대해 많은 논쟁이 있다. 본 연구에서는 이러한 논의를 바탕으로 주식수익률과 채권수익률의 관계에 대해 소파동 상관관계를 통한 분석을 실시하였다. 분석결과 소파동 상관관계는 일정하지는 않지만 전반적으로 양(+)의 값을 갖는 것으로 나타났다. 주식수익률의 분산이 채권수익률 변화분의 분산에 비해 대체로 큰 것으로 나타나, 주식시장이 채권시장보다 충격에 대한 반응이 더욱 크게 나타난다는 것을 의미한다. 주식수익률과 채권수익률의 변화분에 대한 소파동 분석 결과 장기주기에서는 두 변수의 수익률이 함께 움직이는 것으로 나타났다. 이러한 결과는 장기에 주가와 채권수익률이 함께 움직이면서 전략적인 자산배분이 가능하다는 것을 의미한다.

In this paper, we analyzed the economic time series and financial data using the wavelet methods. Some decisions are taken with respect to long-term plans, while other decisions are taken with respect to short-run variations. At different time-scales, a variety of activities of heterogeneous economic agents will interact along the economic variables with different characteristics. Such structures at a different time horizons can be unveiled by the decomposition of time series on a scale-by-scale basis via wavelet. Our main purpose is to present a wavelet methodology for decomposing time series data and to discuss the implications of the wavelet decomposition. The advantage of the wavelet approach is to analyze different time series based on various time-scales. In this paper, we examined the relationship between the stock prices and the bond yields. As theoretical studies argue that this relationship may be either negative or positive, we developed a wavelet correlation analysis model for investigating the relationship. The empirical result shows that the correlation is basically positive although the value is not constant. The variance of changes in stock prices is more volatile in all time-scales than those of changes in the bond yield. From the analysis of the correlation over the wavelet time domain, we found that changes in stock prices and bond yields are more bound in the long scale. The result of the wavelet analysis reveals that changes in stock prices and bond yields does move together in the long scale. The results indicate that a tactical asset allocation may hold in the long-run, because changes in stock price and bond yield move together.

본 논문은 에너지-성장의 탈동조화를 소파동변환(Wavelet Transformation)이라는 새로운 기법을 적용하여 분석하였다. 연구의 목표는 에너지-성장의 탈동조화가 단기, 중기, 장기적으로 나타나는 국가와 나타나지 않는 국가를 식별하고 어떤 요인이 탈동조화의 확률을 높이는지를 분석하는 것이다. 논문의 주요결과는 다음과 같다. 첫째, 에너지소비와 GDP의 소파동분해 후 그랜져 인과성 검정을 수행한 결과 상당수의 국가들에서 각 시간척도별로 인과관계가 바뀌었다. 둘째, 국가들의 에너지-성장 비인과 확률에 유의한 영향을 미치는 변수들은 각 시간척도별로 상이하였다. 에너지효율성은 단기에 에너지-성장 비인과의 확률을 높였으나 중장기에는 유의한 영향을 미치지 못했다. 청정에너지 비중은 중기에만 비인과 확률을 높였고 단기와 장기에는 유의성이 없었다. 에너지순수입은 단기에는 비인과 확률을 높였으나 장기에는 오히려 하락시킴으로써 장단기에 그 영향력이 달랐다. 반면, 제조업비중은 장단기에 모두 비인과 확률을 높였다.

This study investigates the decoupling between energy consumption and economic growth by applying wavelet transformation. The objective of this paper is to divide the 54 countries into that the relationship decouples and does not from the periodic perspectives, and to further identify which explanatory variables determine the decoupling. The main results as follow. First, energy-GDP causality differs by time scale. Second, the variables affecting the decoupling of the energy-GDP relationship are different in the distinct time scale. Energy efficiency increases the decoupling probability in the short-run but is not influential in the medium and long-run. Clean energy ratio increases the decoupling probability only in the medium-run. Net energy import increases the probability in the short-run but decreases it in the long-run. The results suggest us that distinct energy policy is needed in the different time-horizons.

More and more electric vehicles have been used in China, and many charging station has been built in China. Due to the reactive power loss caused by charging devices, the method to measure the power consumed by charging device is critical for power company. Based on impact loads generated by electric vehicles to the grid in fast charging mode, this paper studies the energy measurement method of electric vehicle charging mode. Since each phase current of charging post imbalance, to prevent the sum of each phase current vector of charging posts over the protection current threshold and the charging post breaker misusing in power circuit, preventing leakage circuit breaker current imbalance system is designed. Finally, this paper studies the power consumption detection methods of electric vehicle charging posts. By comparing the fast Fourier transform and wavelet analysis algorithms, an electric energy measurement device of electric vehicle charging machine based on FFT and wavelet analysis are proposed. Combined with fast data processing functions of DSP, the device can quickly and accurately measures electric vehicles charging amounts under different charging modes.

Landslide deformation signal processing is significant for landslide stability analysis. Global Position System (GPS) control networks were built to monitor landslide deformation and acquire landslide displacement time series. It was difficult to predict landslide displacement because of the highly non-linear and non-stationary characteristics contained in displacement time series. A Wavelet Analysis - Radial Basis Function Neural Network (WA-RBFNN) model was proposed to overcome this problem. Firstly, monthly cumulative displacement time series was decomposed into different frequency components using wavelet analysis. Then a RNFNN model was established to forecast each frequency component values. The final prediction results were obtained through the sum of the predictive values of each frequency component. GPS monitoring points ZG325 and ZG326 on Baijiabao landslide in the Three Gorges Reservoir Area were used as study cases. A single RBFNN model was also built as comparison. The experimental results show that GPS control network can monitor landslide deformation accurately and the WA-RBFNN model is of high prediction accuracy. What is more, WA-RBFNN model has better prediction effect than a single RBFNN model.

In order to expand the dynamic range of the GMI sensor in longitudinally excitated amorphous wire and improve its precision, waveforms of the GMI sensor are analyzed on the background of weak magnetic field measurement. Then three features extraction methods are studied in detail. According to the advantages and disadvantages of different methods, an improved method which combines the energy features of the wavelet decomposition and the amplitude features is proposed. First, fit the amplitude change ratio curve respectively with Gaussian function and polynomial function, which not only solves the problem of nonlinearity, but also improves the measurement accuracy. Considering the difference of signals’ in-pulse features at different positions, the ‘db5’ wavelet is introduced to decompose the signals. Then the BP neural network trained by the energy features of the wavelet is used to locate the target’s approximate position, as a result, the problem of multi-value is solved. At last, experiments of target detection in weak magnetic field prove that the method proposed is effective.

Based on the analysis of the defect of traditional model, this paper proposes a new control chart pattern recognition model, which includes Wavelet Analysis (WA), Principal Component Analysis (PCA), Particle Swarm Optimization (PSO) and Support Vector Machine (SVM). WA is good to eliminate noise control chart anomaly pattern recognition of the adverse effect. PCA eliminates the redundant information of data between SVM and reduces the input dimension and computational complexity. PSO algorithm optimizes the parameters of SVM and the establishment of the optimal control chart anomaly pattern classifier can solve the problem optimal parameters of SVM. The simulation results show that the model is feasible, the results are reliable. This algorithm improves the control chart abnormal state average recognition accuracy and be used in the machining process real-time monitoring.

In allusion to the shortcomings of the existing signal analysis method for high-frequency analysis, non-stationary signal analysis and so on, wavelet analysis and grey theory are introduced into the signal analysis, a new signal analysis method based wavelet analysis and grey theory is proposed in this paper. In this method, the wavelet packet is used to nonredundantly, lossless and orthogonally decompose different components of noise signals into different frequency bands with different scales, in order to realize the signal frequency band division with total energy conservation for obtaining the energy feature of each frequency band. Then these energy frequency bands are used to construct the feature vector. And the grey theory is used to analyze the correlation degree between the equipment states and system parameters in order to quickly and accurately determine the position of source. Finally, for a typical signal simulation and analysis, the effectiveness of the signal method is tested and verified.

Oil debris signals are applied to monitor lubricant oil conditions and fault diagnosis of a wind turbine. However, the signals will be influenced by the vibration of transmission mechanism and background noise, hence false alarms and undetected particles will limit sensors ability in examining fine particles in oil debris. This paper presents an approach to enhance the performance of oil debris signals. The de-noising of signals used a wavelet filtering to remove the vibration of a wind turbine, and then set a threshold to remove the background noise which caused by the system of measurement. The effectiveness of this enhanced measurement system is tested by using simulated and experimental signals.

Islanding detection method of active current disturbances in three-phase photovoltaic-to-grid system makes the common connection point (PCC) voltage appear high frequency components on account of load variation, the surge current occurrence and other factors. This high frequency component cause islands misjudgment. Using Db10 wavelet to detect and analysis PCC’s voltage high-frequency components on real-time based on the active current disturbances principle. And selecting an effective wavelet domain values as the voltage harmonic detection amount islanding detection. Islands simulation test is conducted in the case of the inverter output power and load input power to match, the test results show that this method can quickly detect islanding, and effectively prevent pseudo-island phenomenon of false positives.

Fast Fourier Transform plays a very important role in signal analysis, but the Fast Fourier Transform traditional mutation fault fan are unable to analyze the trend of fault features, the beginning and the end, and these signals often contain important information, the fault at the same time, the local signal analysis Fast Fourier Transform of fault are also incapable of action. The method of multi scale wavelet theories and Fast Fourier Transform are combined, make up for the deficiency of the Fast Fourier Transform, and the method is applied to the fault diagnosis of fan, and achieved good results, experiments show that, this method can effectively improve the accuracy of fault diagnosis. Wavelet packet analysis due to the high, the low frequency part of the signal local refinement and retention time domain features of the original signal, which has good time-frequency localization characteristics, it can effectively identify the non-stationary signal, to achieve the purpose of fault diagnosis, get more and more extensive application in the field of fault diagnosis. Signal generating fan running most of the non-stationary signal, the wavelet packet analysis technology is used to diagnose the fault has practical significance.

Reliability is an essential performance in GPS navigation system. Therefore, satellite fault detection is considered as one of the most important functions in GPS receivers. For the declining problem of the ability of fault detection for the traditional GPS satellite fault detection algorithm under the condition of small fault, a new GPS satellite fault detection algorithm based on wavelet analysis is proposed. The raw pseudorange measurements and the positioning data information are transformed by wavelet analysis and the data jumping point can be detected and identified through the different wavelet scales, and the satellite fault could be detected. Two kinds of GPS satellite fault detection methods are given in detail, and the advantages and disadvantages of them are compared. Validated by the real collected raw data from the GPS receiver, the results show that the wavelet analysis method can detect smaller mutations in a sequence of parameters, and the feasibility and effectiveness of applying the wavelet analysis algorithm in satellite fault detection for GPS receiver are verified.

In order to ensure the safe operation of offshore platform, we need response to the platform motion and forecast mooring force. The prediction method based on numerical calculation and model experiment, has certain limitation. A new principle and method of ship’s mooring load measurements based on indirect measurement is presented in order to achieve the short-term and high-precision mooring load prediction, and an algorithm is proposed through which predictions are made by comb the wavelet multi-scale decomposition and reconstruction method with BP neural networks. This paper, by putting a prototype data as learning samples, using the neural network algorithm for forecasting of mooring force, overcomes the traditional B P neural network faults, gets a higher precision. Through comparing the measured data, it demonstrates the feasibility of this method in engineering application.

The unit capacity of propulsion motor is greater than that of generator in all-electric ship electric system, in this high power load changes under the impact of random overload, the generator is extremely easy to have the fault causing system crashes. In this paper, a mathematical model of all-electric ship power system was established, and wavelet analysis was used to extract the feature of heavy load of power grid fault condition based on MATLAB. The simulation results prove that the simulating model of All Electric Ship Power System (AESPS) is reasonable, and by using the method wavelet analysis the feature information can be effectively extracted, which provides the basis for fault diagnosis.

In this paper, we present comparative analysis of scale-invariant feature extraction using different wavelet bases. The main advantage of the wavelet transform is the multi-resolution analysis. Furthermore, wavelets enable localization in both space and frequency domains and high-frequency salient feature detection. Wavelet transforms can use various basis functions. This research aims at comparative analysis of Haar, Daubechies and Gabor wavelets for scale-invariant feature extraction. Experimental results show that Gabor wavelets outperform better than Haar, Daubechies wavelets in the sense of both objective and subjective measures.

There are various types of noises that affect quality of an image such as Salt-and-pepper noise, Poison noise, Gaussian noise, Speckle noise etc. Wavelet is a powerful tool for denoising a variety of signals. Here a White Flower image has been taken for denoising purpose with the help of HAAR Transform. The noisy image is first decomposed into five levels to obtain different frequency bands. Then hard thresholding method is used to remove the noisy coefficients by fixing the optimum thresholding value. In this paper, analysis of a colored image is carried out with four different noises at zero mean that are applied on the image to produce noisy images. Residual image is obtained from the original and noisy image & its statistical parameters such as mean, median, mode, standard deviation, mean absolute deviation, median absolute deviation are calculated. In order to enhance the quality of the noisy images, performance parameters of denoised images must be estimated. The comparison between noisy and denoised image is taken in terms of MSE (mean square error), PSNR (peak signal to noise ratio), RMSE (root mean square error), SNR (signal to noise ratio) and SSIM (structural similarity index).

The paper explains the method of detection of QRS complex from ECG signal using wavelet transform.By using MATLABtool, we can detect QRS complex which further helps us in diagnosis of various disease i.e. related to heart, when QRS complexes were detected then each complex was used to find the peaks of the waves like Q,R& S.

In this paper a novel face recognition approach based on Adaptive Principal Component Analysis (APCA) and de-noised database is produced. The aim of our approach is to overcome PCA disadvantages especially the two limitations of discriminatory power poverty and the computational load complexity, by producing a new adaptive PCA based on single level 2-D discrete wavelet transform using Daubachies filter mode. All face images in ORL database are transformed to JPG file format and are de-noised by Haar wavelet at level 10 of decomposition; the goal is to exhibit the advantage of wavelet over compressed JPG files instead of using origin PGM file format. As a result , our adaptive approach produced good performance in raising the accuracy ratio and reducing both the time and the computation complexities when compared with four other methods represented by standard statistical PCA, Kernel PCA, Gabor PCA and PCA with Back propagation Neural Network (BPNN).

The need for clear ECG signals is increasing to reduce the probability of misdiagnosis regarding heart diseases. An algorithm was designed to diagnose patients with heart diseases using ECG signal analysis so that it can help physicians in the decision-making process. Before analyzing the ECG signal, noise of the low and high frequency components was removed through preprocessing. After preprocessing, pattern analysis detected important features on which diagnosis will be given. Then, the analysis was applied on the pure ECG signal to detect the patient's heart diseases. All feature points were extracted by using the proposed algorithm, called ‘Geometric Detection (GD)’. Results showed that performance was superior to others in standard error of the sample mean and variance. Data from CSE (Common Standards for Electrocardiography) database were used to test each algorithm except for GD, because patients’ ECG data was used to test the GD algorithm. Detection rate of the GD algorithm (se(%)) was 99.1% and we confirmed that the proposed algorithm is superior to the other algorithm in terms of stability and standard error of the sample mean. The result of the performance evaluation showed that the proposed algorithm produced higher accuracy and stability than the other algorithms.

To optimize the number of decomposition layers in wavelet threshold denoising for ultrasonic signals, we propose a self-adaptive algorithm of determining the number of decomposition layers based on singular spectrum analysis and wavelet entropy. First the noise-containing signals are decomposed by discrete wavelet transform. The slope of the singular value spectrum for each layer is estimated. Then the wavelet entropy over the signal subinterval is calculated for each layer. Finally the optimal number of decomposition layer is determined by combining the entropy ratio of detail coefficients to original signal and the slope of the singular value spectrum. The performance of the algorithm is evaluated using signal-to-noise ratio (SNR) and the relative error of the peak value (REPV). Experiment shows that the algorithm can self-adaptively determine the optimal number of decomposition layers and filter out the noise contained in the ultrasonic signals. It not only increases the SNR, but also preserves valuable components of the original signal.