• CN: 11-2187/TH
  • ISSN: 0577-6686

Journal of Mechanical Engineering ›› 2018, Vol. 54 ›› Issue (8): 95-100.doi: 10.3901/JME.2018.08.095

Previous Articles     Next Articles

Translation Invariant Wavelet De-noising of CO2 Gas Shielded Arc Welding Electrical Signal

HUANG Yong, WANG Kehong, ZHOU Xiaoxiao   

  1. School of Material Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094
  • Received:2017-04-25 Revised:2017-11-22 Online:2018-04-20 Published:2018-04-20

Abstract: The electrical signal in the CO2 gas shielded arc welding process contains a large number of random non-stationary noises. De-noising pretreatment is an important part in the analysis of electrical signals. Common signal filtering methods are hardware filter and software filter, in which the wavelet threshold de-noising method is widely used in many wavelet de-noising methods. This method can eliminate the noise in the signal, but it is easy to generate the pseudo-Gibbs phenomenon at the signal discontinuity point. we put forward the method of using translation invariant wavelet to de-noise the welding signal. The method through multiple signal cyclic shift, and soft (hard) wavelet threshold de-noising the shifted signal with soft (hard) wavelet threshold, and then inverse translate the de-noised signal. Finally, the de-noising results are averaged. This method can effectively eliminate the pseudo-Gibbs phenomenon of the traditional soft threshold wavelet de-noising to appear. The result shows that this method can improve the SNR of signal, and the signal de-noising is closer to the true signal. This method has broadly prospects in GMAW electric signal noise reduction processing, and extend the application of wavelet method in welding process.

Key words: CO2 gas shielded arc welding, electrical signal, soft threshold wavelet de-noising, translation invariant

CLC Number: