Analysis of electrochemical noise signals with classical methods and methods based on fractal-like wavelets

Abstract

The corrosion process of low carbon steel and stainless steel in natural aerated sulphuric acid was examined using electrochemical noise (EN). In the paper the comparison of classical Fourier/statistical methods for analysis and characterisation of EN were compared by wavelet based methods. The approximation properties of discrete wavelet transform (DWT) using smooth and fractal-like wavelets were compared for EN-signals. There were found very little differences in approximation rates for both corrosion processes EN-signals. It was shown that DWT enable additional time-frequency characterisation of broadband EN-signals.