Abstract : The research work is dealing with variable speed wind turbines modelling and control design, in order to achieve the objectives of maximizing the extracted energy from the wind, below the rated power area in the one hand and in the other hand regulating the electric power production, above the rated power area, while reducing mechanical transient loads.
Thus far, wind turbines control is achieved by using classical regulators by means of PI or PID. The related performance are acceptable for low turbulent wind speed, however, they dramatically decrease during fast wind speed variations. So then, due to the strong non-linearity of the wind turbine aerodynamics, its dynamic aspect and the turbulent nature of the wind, advanced control techniques are being imperative to ensure more efficient wind power system operation.
For this purpose, we have studied various control strategies from linear to nonlinear based approaches and for which the advantages and drawbacks have been outlined with regards to the required performance. Some of the controllers that we have developed, herein appear for the first time in the relevant domain, the remaining others are an adaptation of well know controllers to the adopted wind turbine models. As matter of fact, we have derived two wind turbine models as well as a wind speed estimator. Indeed, the estimator allows obtaining the effective wind speed which cannot be measured, since the wind profile around the rotor is variable in time and space.
As results, it has been shown that single input control by means of pitch angle or generator control cannot succeed to simultaneously drive the electric power output regulation and the rotor speed reference tracking. So then, our idea is to combine nonlinear dynamic state feedback torque control and pitch linear based control which turns out to be the best strategy. In addition, the validation of the controllers performance, using a high turbulence wind speed profile, has been performed through wind turbine simulators provided by NREL (National Renewable Energy Laboratory, Golden, CO), has confirmed the theoretical results and has led to quite satisfactory conclusions in terms of energy capture optimization, power regulation and disturbances strong rejection as well.