Artificial intelligence-based blade element momentum method for wind turbine systems

Abstract

The output aerodynamic power from a wind turbine is estimated through a classical c1 - c6 formulae in most of the research works especially when it is considered for the generation of electrical power. This approach sometimes may not be useful where the actual aerodynamic power with better accuracy is required. This paper investigates the blade element momentum (BEM) method in-depth with the impact of wind speed, turbine speed and air-foil geometry. An artificial intelligence model (AIM) of BEM for its use in simulation has also been proposed in this paper. AIM helps to reduce the computational time significantly since the BEM when run in whole takes a lot of time during simulation. A neural network has been made and trained with the data obtained from the BEM method. Further, the turbine power resulted from the BEM approach through AIM has been used for the generation of the electrical power with its maximum power tracking. The simulation has been performed on NREL's 5-MW test wind turbine.