Multiobjetive H2/H control of a pitch regulated wind turbine for mechanical load reduction.

Abstract

The paper deals with the design of a control system for a variable-speed pitch-regulated wind turbine in the operating region corresponding to wind speeds above rated wind speed. The control objectives are mostly to regulate the plant output power and to reduce the mechanical fatigue of the plant components. The controller is designed from a non linear model of the system which takes into account the blades, shaft and tower flexibilities. A Linear Matrix Inequalities (LMI) formulation of the control problem permits to obtain a Linear Parameter Varying (LPV) controller optimizing, through H 2 /H ∞ norm minimization, a criteria close to mechanical fatigue affecting the plant. The proposed controller performances are then compared in simulation with those of a gain scheduling Linear Quadratic Gaussian (LQG) controller and with a gain scheduling PI controller, and is seen to be more efficient.

Wind Tunnel Testing of Subspace Predictive Repetitive Control for Variable Pitch Wind Turbines

“…In recent literature, multiinput multioutput (MIMO) control techniques such as linear-quadratic (LQ) [14], fuzzy logic [15], model predictive control (MPC) [16], linear parameter-varying (LPV) [17], and H 2 -H ∞ [18] controllers have been tested for wind turbine power production. Conventional IPC is typically based on designing a PI controller for different operating conditions and for each load peak to be targeted.…”

A tutorial on the dynamics and control of wind turbines and wind farms

“…Assuming more sensors/measurements are available for feedback, multi-input, multi-output (MIMO) individual blade pitch controllers can be designed for increased performance [23], [24]. Additional discussion on MIMO pitch feedback control is provided in [25].…”

“…The coupling across degrees of freedom becomes even more pronounced in floating offshore wind turbines; Section VIII will discuss additional issues in offshore wind turbines. A unified multi-input, multi-output (MIMO) framework for individual blade pitch control can be beneficial and has been shown to achieve significant load reductions [23], [24].…”

Representation of fatigue for wind turbine control

“…reductions in loading are achieved by controlling the pitch of each blade independently; the damage of different control strategies is assessed by EDL. One could say that the standard to verify if a control strategy actually reduces long‐term loads is either EDL or RFC, see for example Lescher et al ., Nourdine et al ., Soltani et al ., Madsen et al . and Mirzaei et al .…”