EN TRANSIENT ANALYSIS THROUGH WAVELETS

摘要

Statistical and spectral analysis methods are often used to study electrochemical noise time records. However, current and voltage noise signals show characteristic transients in some systems. In these cases, the analysis of the number, shape and distribution of the transients may provide information more valuable than the only mean value amplitude and frequency distribution of the fluctuations. This work proposes an algorithm to automatically detect the existence of transients and even provide a comparative characterization of their size and scale. This algorithm is based a mathematical tool -wavelet transform-, which enables a simultaneous analysis of signals in both time and scale. Thus, this work starts with a brief introduction on wavelet transform to then presents the algorithm and applies it to a current time record and its corresponding voltage signal to illustrate the methodology.