The effect of signal-to-noise ratio value on the error in measuring acoustic emission parameters: statistical assessment

Modern acoustic emission diagnostic systems and complexes are a sensitive tool for detecting developing defects at an early stage when monitoring the technical condition of objects under operational loads. A significant limitation of the application acoustic emission method is the difficulty in isolating signals against the background of acoustic and electromagnetic interference. The effect of interference during acoustic emission recording significantly complicates the interpretation of parameters that characterize the technical condition of the test object. To increase the value signalto- noise ratio and increase the reliability of the results of acoustic emission testing in the quantitative assessment of parameters, filtering methods are used. The subject of this study is the study of the effect of signal-to-noise ratio value on the measurement error acoustic emission parameters formatted during noise compensation using the polynomial filtering method. The basis of the statistical model characterizing the effect of signal-to-noise ratio value on the measurement error acoustic emission parameters is based on the machine learning method — linear regression. The dependence of the measurement error on the signal-to-noise ratio value was approximated by the least-squares method and visualized using a scattergram. It was found that when using the Butterworth filter, the relative measurement error acoustic emission parameters do not exceed 3 %, which are orders of magnitude lower than the values obtained for the Bessel filter and Daubechies mother functions 8 based on wavelet filter. A high inverse non-random correlation was established (r > 0.9), due to a decrease in the values of the relative measurement error emission parameters and an increase in the signal-to-noise ratio value. The developed statistical model describes the effect of the signal-to-noise ratio value on the value relative error in estimating the acoustic emission parameters. The adequacy of the developed model was confirmed by calculating the coefficient of determination and checking its statistical significance. It is shown that the use of Butterworth filter to compensate for interference significantly increases the information content of the results of measurements of acoustic emission parameters. The developed statistical model can be used in the development of new or improvement of existing complexes and systems for processing acoustic emission data to improve the reliability of the results of acoustic testing.