Optimal Design of a Grid-Connected-to-Rotor Type Doubly Fed Induction Generator for Wind Turbine Systems

This paper performs a shape optimization of a grid-connected-to-the-rotor type doubly fed induction generator (DFIG) in which the rotor winding is connected to the grid instead of the stator winding. To analyze the characteristics of a DFIG accurately, a time step 2D-FEA is carried out, which is coupled with Simplorer. To maximize the torque per weight as well as to maintain the efficiency, the Kriging model based on latin hypercube sampling and genetic algorithm are utilized. As an optimal design result, the torque per weight on the optimal model has been increased. Finally, the analysis and the optimal design results are confirmed by the experimental results.