Acoustic emission source locating in two-layer plate using wavelet packet decomposition and wavelet-based optimized residual complexity

Abstract

SummaryHealth monitoring based on acoustic emission principle needs precise time delay estimation in two-layered plate-type structures. In this paper, the theories of wavelet packet decomposition, wavelet-based optimized residual complexity (WORC), and frequency-varying velocities were used to acoustic emission source locating. A rectangular array of the four sensors was used to locate acoustic emission source. By wavelet packet decomposition, specific packets with frequency range of 0-250 kHz were selected for more signal processing. Then WORC of specific packets of captured signals was calculated as a similarity measure technique. The time delay was estimated when WORC function reached the minimum value.The group velocity was obtained using dispersive curves. The experiments were carried out, and the results of locating error showed the high precision of the proposed algorithm. | INTRODUCTIONPrecise and robust defect source locating is one of the most important issues in structure health monitoring. The acoustic emission (AE) technique is one of the real-time and in-service methods for this purpose. The main advantages of approaches based on AE are their simplicity and high sensitivity. Mostafapour et al. [1] used the theories of wavelet transform and cross-time frequency spectrum to locate AE source with frequency-varying wave velocity in plate-type structures. They used a rectangular array of the four sensors on the plate. Yang et al. [2] proposed an AE analysis method based on multiple signal classification method to calculate the direction of arrival of wave signal in plates. They estimated these directions using the multiple signal classification, and the time delay of the wave was gained using the continuous wavelet transform. Jumaili et al. [3] presented an improved automatic delta T mapping technique using a clustering algorithm to automatically identify and select the highly correlated events at each grid point. They used "minimum difference approach" to determine AE source location. Ding et al. [4] developed a new waveform analysis to estimate AE wave arrival times using wavelet transform. A variety of techniques were compared in the proposed method such as threshold crossing, cross correlation, and wavelet packet transform, and a method based on wavelet decomposition was recommended as the most consistent and accurate method for determining arrival time. Scholey et al. [5] published a method where a map was generated on the plate with arrival time differences for each pair of sensors. The time difference was calculated using sensor distances and wave velocity dependency on the fiber orientation. This technique is named as "best match point search method." Necessity to a large number of training points is an important

State-of-the-Art Review on the Acoustic Emission Source Localization Techniques

“…Results obtained at multiple frequencies were then blended by a data fusion scheme to maximize the accuracy and minimize the uncertainty regarding the final location. Mostafapour and Davoodi [87] used a rectangular array of four sensors. The approach proposed by Perelli et al [96] succeeded to overcome the issues related to the detection of arrival time using conventional threshold methods.…”

Discrete wavelet analysis of acoustic emission precursory characteristics of confining pressure failure in Coal and rock triaxial unloading

“…The research shows that acoustic emission parameters such as acoustic ringing count rate and energy rate can better describe the damage and rupture of rock samples during the test, and the damage and rupture of rock samples can be predicted by acoustic emission parameters [13][14][15][16][17][18] . However, as a discrete non-stationary signal, acoustic emission parameters are di cult to be effectively processed by conventional methods, while wavelet analysis is an effective method for signal processing, especially for such non-stationary signals [19][20][21][22][23].…”

An instantaneous frequency analysis method of stay cables

“…22 WT uses scaled and shifted wavelets and functions orthogonality to simultaneously extract time-localized components; WT can adjust the window functions so that the window size will widen and narrow according to low and high frequencies. 8,23,24 However, the successfully extracted instantaneous frequency (IF) of WT greatly depends on the selection of the wavelet basis function and the discretization of scales; WT suffers from limitations posed by the uncertainty principle.…”