Recognition of signals for stress concentration zone in metal magnetic memory tests

Metal magnetic memory (MMM) technique is an effective method to find stress concentration zone, which provides early diagnosis for the ferromagnetic material. Metal magnetic memory signal is weak and easy to be influenced by noise and disturbance. Thereby the stress concentration zone is difficult to be detected. The two methods are presented for non-stationary leakage signals based on original feature gradate. One is a combination of the peak-peak and grads, the other is mean energy method, with the methods eliminate influence of noise and the stress concentration identification is improved. Based on the relativity for wavelet coefficient square, the signal singularity detection theory of wavelet transform was employed to determine metal magnetic memory. Experiment shows that the method is valid for leakage signal.