Papers by Keyword: Wavelet Analysis

Abstract: Optical profiler is employed to measure the surface topography of H-13 die steel machined by ball end cutter. By using 2D discret wavelet decomposition, the topographies of different frequency bands are obtained and the influence of cutter posture on varying frequncy bands topography is studied. The results show that cutter posture has a great influence on the whole frequency band roughness. The changing trend of roughness is roughly the same under different cutter postures which increase firstly then decrease with the decrease in frequency. A surface with a small high frequency roughness may have a large median frequency roughness.

Abstract: Vibration has a great influence on the cutting process and it can be detected by many signals. In this work, a set of experiments are conducted on Mikron UCP810 DURO high speed milling center with Fraisa carbide end mill for milling 2A12 aluminum workpiece, an acoustic method is used to detect the milling signals. The captured sound signals are analyzed using Matlab Daubechies5 wavelets with six levels of decomposition, the detail and approximation of the sound signal components are obtained. The analysis results demonstrate the relationship between the signal and the vibration.

Abstract: Optical profiler is employed to acquire topography height data of ball-end milled die steel surface under different spindle speeds ranging from 2000rpm to 12000rpm with lead angle of 20° and tilt angle of-10°. By multi-scale wavelet analysis, measured height data are decomposed and then been reconstructed, meanwhile 3D topography and 3D roughness in different frequency bands are obtained. The results show that the changing trend of roughness with frequency band under different spindle speeds is not the same. In the high frequency bands, roughness has a tendency to increase with the increasing spindle speed. In the median frequency band, the roughness of the surface machined under low spindle speed 2000 rpm is the largest and the roughness of the surface machined under high spindle speed 12000 rpm is the lowest. In the low frequency bands, the roughness of the surface machined under low spindle speed 2000rpm is much larger than those obtained under other spindle speeds, and with the increasing spindle speed, the changing trend of roughness increases firstly then decreases.

Abstract: The present paper studies the indoor radon variations in two similar dwellings, one of them occupied and the other unoccupied, from the same residential building. In particular, the research study was carried out in two new dwellings in Madrid. Radon concentration and ambient parameters were measured three-hourly from 11 May to 12 December 2014. Solid state detectors, Radon Scout, were used for the radon monitoring. Simultaneously, two atmospheric variables were acquired from NOAA Air Research Laboratory. The results show that radon level is slightly higher in the unoccupied dwelling than in the other one. From the analysis developed in this study, it is found that a specific seasonal pattern exists in the indoor radon concentration. Besides, the anthropogenic influence is also analysed. Nearly periodical patterns could be observed in specific periods whether dwelling is occupied or not.

Abstract: ESOUNDMAPS is an ongoing research program that aims in developing a wireless audio sensor network (WASN) and deploying it at the surrounding environmental area of the Technological Institute of Piraeus in Attica, Greece. The proposed WASN will be used for the environmental monitoring of the area and aims to be used in the future for assessing the impact of human generated noise to the wildlife. Collected sound samples will be forwarded by the sensor nodes to a central server where they will be automatically evaluated with respect to their identity; environmental sound maps will be generated based on the evaluation of these sound samples. High-level sound classification is defined herein as the act of classifying a sound sample to three broad categories namely anthropogenic, biophysical (other than human) and geophysical sounds. In this paper we present an integrated platform that includes sound sample denoising using wavelets, feature extraction from sound samples and Gaussian mixture modeling of these features, and a powerful two-layer neural network classifier for the automated high-level classification of incoming sound samples. Classification results, obtained using digital sound samples, exhibit outstanding classification accuracy (sometimes reaching or exceeding 98% correct vs. incorrect estimates), thus demonstrating the feasibility of the proposed approach in realistic environments.

Abstract: Voltage sag is the major power quality problem and receives wide attention. Although wavelet analysis works well for detecting voltage sag features, the existence of noise can reduce the advantages of wavelet method or even make it ineffective. To solve the problem, the paper uses multi-scale wavelet information by multiplying the results of several scales, and then searches the local maxima from the product to find the transition moment of voltage sag. The proposed method can suppress the noise and improve the accuracy for detecting voltage sag features. Simulation result validates the effectiveness of the proposed method.

Abstract: Through the Merlot wavelet periodic analysis and Mann-Kendall mutation analysis about the maximum wind speed of 14 weather station in Hexi corridor over the past 41 years,the evolution characteristic of the maximum wind speed can be gotten:the Hexi corridor has about 28-29 years main period and 8-12 years secondary period,presenting sinusoidal curve,and the transformation energy cycled from small to big to small to big;After 1980,there were abrupt changes and decline in Anxi,Jiuquan,Zhangye and Wuwei.The Hexi corridor is in the wind speed environment of same climate conditions,and the inter-annual change feature becomes balanced; but in internal local areas,due to the influence of environment, terrain and urban heat island effect,the secondary period is different,presenting abrupt changes and decreasing tendency in 1980s.

Abstract: In this paper, the defect sheet was captured coincidentally. According to the defective product’s characteristics, we suspected to be caused by the vertical vibration of the roll. When the rolling speed reached a certain value, the vibration of the fourth stand can be feel. The experiment of the vibration data collection was taken to compare the vibration parameters of rolling operating side with those of drive side by wavelet analysis. The result states that the abnormal vibration signal features can be extracted in a special frequency segment of wavelet decomposition, and the vibration frequency to the roll is confirmed which appeared product defects.

Abstract: Due to the influences of climate change and human activities, the water and sediment flux of the Yellow River are certainly changing. This paper selects monthly time series of runoff and sediment flux from 1950 to 2009 for study at Lijin station, in lower Yellow River. A widely used identification method, wavelet analysis, is applied for recognizing changing point and cycle of the runoff and sediment respectively in multi-scale of annual, flood season and non-flood season. The results indicate that there are two significant changing points in 1985, 2002 year, and cycle recognized results are different in multi-scale as well as with different hydrology factors.

Abstract: Due to IGBT open-circuit fault of CRH2 EMU’s traction inverter, a method of its fault diagnosis with the three-phase current signals as detection objects is conducted. By applying the wavelet analysis, three-phase current signals are decomposed for four times. With the coefficients of each layer obtained, the energy values of layers are calculated as well as the vectors corresponding to failure modes. According to the vectors regarded as input and the expected output, a BP neural network is established. Through training the network, the parameters of network can be defined. In addition, to test and evaluate the performance of network, certain noise is added to the three-phase current signals. Simulation results show it is feasible for the fault diagnosis of traction inverter.