Optimal Harmonic Estimation Using an IGHS Algorithm

This paper presents an improved global-best harmony search (IGHS) method to estimate the harmonics in power systems. It utilizes IGHS to estimate the harmonic components of a distorted signal along with adaptive noise. The harmonic estimation problem is linear in amplitudes and nonlinear in phases and frequencies. IGHS is used to estimate phases and frequencies, whereas the least square (LS) method is used to estimate the amplitudes. The improvements in error, as well as the running time compared with the conventional discrete Fourier transform, genetic algorithm method, particle swarm optimizer with passive congregation (PSOPC) and Fuzzy Bacterial Foraging (FBF) is explained in this paper. The results show that new method is able to estimate integral harmonics, inter harmonics and sub-harmonics even in the presence of fundamental frequency deviation and decaying DC component with quite acceptable performance.