Improvement of wind farm power quality by means of ANN-contorlled hybrid filter

Converters that are specifically designed for use in wind turbines may inject considerable amount of harmonics into their connected distribution systems that could exceed its tolerance limit determined by IEEE 1547 Std. In this paper a hybrid filter constructed of a shunt active filter series with passive filters is used for power quality and a neural network based approach is proposed to control the hybrid filter. This method allows for accurate calculation of cancellation references filter operating under steady-state and transient condition. The most important advantage of the suggested model in comparison with previous works is the removal of the low-pass filter and therefore reducing the processing time. The method is simulated with Matlab/Simulink and the simulation results have shown the validity of the proposed neural network based scheme.