Best Wavelet Packet Power Measurement Based on Energy Threshold Entropy

Abstract

Firstly, a sub-band energy threshold entropy was proposed as information measurement cost function. Then a novel best wavelet packet power measurement algorithm was proposed, which can decompose power signals into sub-bands adaptively according to the energy distribution, meanwhile, this algorithm can perform the decomposition of power energy in one wavelet packet process. The simulation using db40 wavelets shows that: the measurement relative errors of root mean square (RMS) and the active power are less than 6×10-5 in the total band, and the quoted errors of active power are less than 8.2×10-5 in the sub-bands.