Electrical Loss Minimization Technique for Wind Generators based on a Comprehensive Dynamic Modelling of Induction Machines

This paper proposes a novel model-based electrical loss minimization technique (ELMT), whose main original contribution lies in the overall power loss function which has been derived from a comprehensive dynamic space-vector model of the Induction Machine (IM) including the iron losses, expressed in the rotor flux oriented reference frame. Such a loss formulation, obtained from the IM input-output power balance, is more general and accurate than the others in the literature; consequently, the expression of the optimal efficiency reference flux to be given to the FOC control system is more general and accurate too. The proposed ELMT has been integrated into an IM-based wind generation system including a previously developed Maximum Power Point Tracking (MPPT) based on a Growing Neural Gas (GNG) artificial neural network. The obtained results show that the new formulation of the overall power losses of the IM leads to an increase of the IM efficiency with respect to the classic loss equation proposed in the scientific literature. The integration of the proposed ELMT in a real wind generation system leads to an increase of the active power injected into the grid ranging from 33% at high wind speeds up to 200% at low wind speeds.