Застосування фрактального аналізу для діагностики несправностей в аналогових електронних схемах

Abstract

This paper presents a fault diagnosis method for analog electronic circuits combining Ensemble Empirical Mode Decomposition (EEMD) with Multifractal Detrended Fluctuation Analysis (MF-DFA). Unlike FFT- and wavelet-based techniques, it detects changes in nonlinear circuit dynamics before clear spectral deviations appear, without requiring signal stationarity. Extracted multifractal parameters form a compact feature vector for automated fault classification. Recent studies suggest that EEMD with multifractal analysis enhances soft fault detection and offers a promising approach for early-stage diagnostics.