Robust neural network scheme for generator side converter of doubly fed induction generator

Abstract

Currently, variable speed wind turbine (WT) generators are extensively used in wind energy conversion system (WECS). Variable speed constant frequency (VSCF) generators assist in better exploitation of wind energy and enhancement of WECS efficiency. However, control action is required to provide stable and reliable operation of variable speed WECS. This paper emphasizes on doubly fed induction generator (DFIG) based WECS. Therefore, a robust neural network (RNN) is proposed for DFIG to access a direct power control scheme for generator side converter (GSC), compared with classical vector control (VC). The DFIG d-q modeling is performed in MATLAB/Simulink using synchronous reference frame. Moreover, a field oriented control (FOC) scheme is employed to control the active and reactive power of GSC. The proposed control model is built in MATLAB and accuracy of results verify the conclusion. Finally, the results of RNN control scheme are analytically and critically compared with conventional proportional integral (PI) control scheme.