Real-time pre-processing of the pulsed eddy current signal from continuous casting slabs.

Both the software and hardware of a pulsed eddy current (PEC) real-time testing device are designed to pre-process the noise from continuous casting slabs (CCS) during in situ defect detection. An algorithm based on a sliding window with a fixed length is developed in the process of denoising. The root mean square error (RMSE) and the signal-to- noise ratio (SNR) of the denoised PEC response are given with the combination of Daubechies4, Symlet4 and three threshold denoising algorithms. The results of the simulation and experiment indicate that the best real-time denoising condition is with the wavelet transform of three or four layers. Herein, the SNR obtained is above 19 dB and the average throughput rate is 51 kHz when the moving speed of the casting slabs is about 2 m/s. Our system highlights real-time processing, stability, high accuracy and strong anti-interference, so it is suitable for automatic testing in a production situation. [ABSTRACT FROM AUTHOR]