论文部分内容阅读
It is difficult to interpret the ultrasonic testing signals as the echoes from different features overlap.In this paper,a pulse compression technique,Wiener filtering followed by autoregressive spectral extrapolation,was used in signal processing of defect reflection to increase the time resolution.The influence of the autoregressive order and width of the frequency window in autoregressive spectral extrapolation were studied in detail and the best parameters were optimized.Then the technique was applied to quantitative analysis of defects in the plate and identification of the closely-spaced defects.It has been shown that the width of defect in plate can be estimated with high accuracy and the closely-spaced defects can be well identified in both simulation and experiment by using this technique.