Loss minimization control of DFIG variable speed wind turbines based on neural network

Different electromechanical models have suggested for generation of electric energy from wind, which are mainly consisting of variable and fixed speed types. Generally, variable speed models are preferred because of their reduction torque undulation; get more energy and high efficiency ratio. However, the cost of power electronic converters and their related control equipments are very high. Doublyfed induction generators present operation of variable speed with better specialty and lower cost. In this paper, is suggested the models of a wind power system, with the aim loss minimization control in order to improvement system operation. A wind power system equipped with DFIG and a VCC that connected between the rotor and the grid, are presented. PI controller and the NARX NN controller’s performance results with voltage and with three phase-to-earth fault are compared. The results show that better dynamic characteristics can be obtained using NN controllers.