Multi-feature extraction for power system disturbances by wavelet transform and Fractal analysis

This paper proposes a multi-feature extraction algorithm aiming to recognize disturbances by analyzing synchrophasor data from wide-area monitoring systems. The novel algorithm utilizes wavelet transform (WT) to extract preliminary features from synchrophasors. Two types of features are created by organizing scaling coefficients and wavelet coefficients from WT. The multi-features are compressed by Fractal analysis (FA) so that a simple form of features is created for application convenience. A Benchmark 5 bus system is used to verify the proposed algorithm. The results indicate that the algorithm is effective to recognize the type and location of power system disturbances.