Repetitive model predictive control for load alleviation on a research wind turbine using trailing edge flaps

Abstract

SummaryThis paper presents the results of an advanced control strategy that employs active trailing edge flaps to reduce the fatigue loads of an experimental wind turbine. The strategy, called repetitive model predictive control, is a multiple‐input multiple‐output controller that aims at the alleviation of out‐of‐plane blade root bending moments. The strategy incorporates the control commands, output errors, and state deviation from the previous rotation. This way, the time lag in the strain sensor input due to the blade inertia is compensated. Additionally, a strategy to limit the computational costs is presented. The load alleviation performance is evaluated at different yaw cases and compared with different individual flap control strategies. The repetitive model predictive control is able to reduce the fatigue loads by up to 23% compared with the better performing individual flap control strategy. This improvement in load reduction is accompanied by an increase in flap travel of up to 7% compared with the individual flap control strategies.