Response Surface Based Optimization of Aerodynamic Performance of a Horizontal Axis Wind Turbine

Abstract

Risoe wind turbine is a 100 kW horizontal axis wind turbine developed by Denmark Technical University National Laboratory for Sustainable Energy to be used for field testing purposes. In this paper, the chord and twist distributions of Risoe wind turbine blades are optimized so that the power production of the wind turbine is maximized. To constrain the mass of the wind turbine blade, the projected surface area of the blade is constrained by its original value. 18 design variables are used to describe the chord and the twist distributions of the blades. The upper and lower bounds of the design variables are taken as 130% and 70% of the original values of the design variables, respectively. The fmincon function of MATLAB based on sequential quadratic programming algorithm is used to solve the optimization problem. A response surface based approach to reduce the computational cost. The optimization results show that the power generated by the wind turbine is increased by 20% through adjusting the chord and twist distributions. It is also found that the response surface based optimization and direct optimization yield nearly the same results, whereas the response surfaces reduce the computational cost by 85%.