Fuzzy adaptive pitch controller of a wind turbine

In case of wind turbine induction generators which are directly connected to grid a control problem arises when the wind speed increases above the rated wind speed, some measure needs to be taken to limit the aerodynamic torque of the wind turbine in order to keep the output power at its rated value. For this purpose various controlling variables may be chosen, like generator speed, generator power and wind speed. There are number of self-tuning controllers, Tan's controller is a model reference adaptive controller (MRAC) and has been tested on various non-linear plants and has proved to be robust with tight control performance. Here we have tried to test the proposed self-learning neuro-fuzzy controller by Tan [2] for the pitch angle control of wind turbine using simulink. The self-learning neuro-fuzzy control strategy has the potential when the system contains strong non-linearity, such as wind turbulence is strong. The self-learning neuro-fuzzy model will try to develop the inverse plant model of the system and will use that to generate the required control action to keep the output at its rated value. In order to carry out this comparison wind turbine induction generator (WTIG) from Simulink distributed resources is used. Comparison of the self-learning neuro-fuzzy control and PI controller has been carried out in different wind profiles and overall results show that self-learning neuro-fuzzy controller can give better results in presence of strong wind disturbances.