Performance improvement of rotor current controller in doubly fed induction generation wind turbine with artificial intelligence methods
Omar Alruwaili, Moayed Mohamed
Abstract
This article deals with improving rotor control through the use of artificial intelligence (AI). Two AI methods are used, studied, and compared with the classic PI controller, namely an artificial neural network (ANN) based method and a deep learning (DL) based advanced method. We show that shallow ANN and DL methods improve the real-time behavior of the rotor by improving the initial time of the output waves and by allowing control on the rotor side as well as on the stator side. The results show that both methods bring performance improvement compared to the classic PI control. This finding reveals that in contrast to the static PI controller, ANN gives satisfactory results related to input and output waves, meanwhile DL shows very good results in real-time and outputs in current, voltage, and power.