Harmonic Spectrum Estimation and Analysis of the Voltage at the PCC of the Distribution Network Connected to Double Fed Induction Generators (DFIG) Based on Nonparametric Algorithm of Multi-Taper

Among the conventional technologies of the wind turbine generator, the use of double fed induction generators (DFIG) is more common, because of the control capability, ability to supply power with constant voltage and frequency, and more appropriate setup costs. The back to back converters of these turbines cannot completely neutralize the electrical and mechanical fluctuations, thus, the fluctuations caused by changes above 25% in the nominal speed range of rotor, network transient errors and voltage imbalance were effective in the turbine output waveform, and cause the frequency offset, flicker and injection of harmonic and inter-harmonic components. Thus, the harmonic spectrum analysis of the distribution networks in the presence of DGs using the conventional methods will be associated with lower errors and accuracy. There are various methods for the harmonic spectrum estimation and analysis of the voltage and current parameters. The use of notch, Kalman, adaptive, intelligent neural networks filters is a sample of harmonic estimation techniques. In addition to the above techniques, the use of parametric, non-parametric and subspace algorithms is also considered for harmonic analysis. In this article, the harmonic spectrum analysis at the PCC point (PCC) voltage of the distribution network to a DFIG wind unit was evaluated using the non-parametric algorithm of multi-taper in the voltage balance and imbalance conditions. The results of simulation have been compared using three methods of FFT, Period gram, Welch's. The results show that the multi-taper algorithm has high accuracy and very low error in the same conditions.