DWT-PSD extraction feature for defect diagnosis of small wind generator

In this paper, the ability to detect broken rotor bar (BRB) defects in a small renewable energy system (based on a squirrel cage induction generator (SCIG)) by the digital signal processing of captured phase currents, is presented. The new approach proposed in this study is a combination of two techniques. The first technique is a discrete wavelet transform (DWT) by the decomposition of the phase current signal in multilevel frequency bands. This is performed with the analysis of some selected approximations and/or details, which contain both the lower and upper sideband components presenting the characteristic frequency of the BRB fault. The second technique is power spectral density (PSD) analysis. This approach provides the ability to optimize the diagnosis of rotor defects in electrical generators. The results obtained by the proposed DWT-PSD approach are proved and improved by comparing them with the results of the PSD analysis, obtained from the original phase current signal delivered by the 5.7-kW squirrel cage induction generator, based on a small wind energy conversion system.