Search

The aim of the present research work has been to design an optimal MIMO LQG controller to reduce the drive-train, blades and tower mechanical stresses of a wind turbine (WT), and at the same time, to involve the WT in the grid primary frequency regulation when it is operating in full load (FL) zone. To verify the effectiveness of the proposed controller, the achieved results are compared to those obtained by a base-line controller based on a PI regulator. Simulation results show that thanks to these controllers, WT can effectively contribute to the grid frequency regulation, tracking tightly the generator power reference which depends on that frequency. Compared with the base-line controller, the LQG controller significantly reduces the mechanical stresses of the WT?s most costly components. [ABSTRACT FROM AUTHOR]

Abstract: The objectives of this work have been to design and analyze a discrete Linear Quadratic Gaussian (LQG) controller which contributes to the primary frequency control of the grid and to the reduction of the drive-train fatigue loads of a Wind Turbine (WT). The WT rotational speed and electrical power are controlled using the generator torque and the pitch angle as control variables. The rotational speed and electrical power references are generated in a higher control level depending on the wind speed and a frequency droop. The controller has been designed in Matlab® and has been simulated in Simulink®, in some linear simulation models corresponding to different operating points. These linear models have been obtained from a 5 MW WT model implemented in the commercial wind turbine simulation tool Bladed. The results validate the efficiency of the designed controller, in comparison with another discrete LQG controller which has not been designed to reduce drive-train fatigue loads. [Copyright &y& Elsevier]