Associations between predictors of acoustic emission signal processing system: accuracy and noise resistance analysis

Abstract

In this article presents the results of analysis of the association between the predictors of root-mean square error, signal-to-noise ratio and correlation coefficient, which characterize the functioning methods of digital acoustic emission signals processing. The analysis and evaluation of predictors developed at the output of digital high-pass filters Butterworth, Bessel and Daubechies wavelet are carried out. For first time, a significant association between the predictors of root-mean square error, the signal-to-noise ratio and the correlation coefficient for the Butterworth digital filter with a correlation coefficient R > 0.9, associated with a simultaneous increase in the noise resistance and processing accuracy of the acoustic emission signal in comparison with the Bessel filters and the Daubechies wavelet, was determined and established. The presence of an association between these predictors is confirmed by the calculation of the correlation coefficient and determination, testing of statistical significance. The identified association of predictors can be considered when developing a classification scheme for filtering methods in terms of accuracy and noise resistance.