LIDAR based multivariable H∞ feedforward control for load reduction in wind turbines

Abstract

Abstract The performance of LIDAR assisted control is highly affected by the quality of the wind measurement. Consequently, it seems convenient to include some kind of uncertainty model during the design of the LIDAR-based feedforward controllers. Robust control methodologies provide a great tool to deal with uncertain systems. In this paper, three H∞ feedforward controllers have been designed based on linear frequency-domain models of the system. Because the control problem is multivariable and multiobjective, the controllers show different combinations of control actions -pitch and generator torque- and control objectives -generator speed regulation and tower base load reduction-. The controllers have been tested in a nonlinear simulation using the aerolastic code OpenFAST, a turbulent wind field and the onshore NREL 5MW wind turbine model.