Abstract
Real time estimation of power system harmonics requires the proper choice of filter structure and algorithm for coefficient adjustment. Though LMS (Least Mean Square) adaptive filter has an elegant and simple structure; the performance degrades in low SNR and non-stationary condition. On the other hand LMF (Least Mean Fourth) adaptive filter provides better estimation but the performance is limited due to high computational complexity. Thus LMF is not a suitable choice for real time estimation of harmonics. In this paper; combined LMS/F adaptive filter is modified using Volterra expansion of input samples up to second order to form VLMS/F filter. This filter offers advantage in terms of estimation accuracy and computational complexity in a low SNR condition. Estimation results of VLMS/F are compared with other variants of LMS for harmonic parameters and decaying DC.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
