Benefit of wind turbine preview control as a function of measurement coherence and preview time

Abstract

Wind turbine control is typically feedback only, relying on measurements from a generator-speed sensor, and sometimes strain gauges and accelerometers. Recently, lidar and other technologies that provide a preview measurement of the incoming wind speed are also becoming available. This raises a question: How much benefit do these preview measurements provide? In this paper, we focus on answering this question for wind turbine collective blade pitch control in above-rated wind speeds, with the objective of minimizing a cost function that includes both generator-speed error and blade pitch actuation. We assume the International Electrotechnical Commission (IEC) Kaimal wind spectrum and the National Renewable Energy Laboratory (NREL) 5-MW turbine model, linearized at various operating points (wind speeds). We then use ℌ2 synthesis to design an optimal combined feedforward/feedback controller that depends on both the amount of preview time available in the wind speed measurement, and the coherence between the wind measurement and the wind that is actually felt by the turbine. Finally, we show how the resulting closed-loop cost decreases as a function of measurement quality (coherence bandwidth) and preview time.We include the special case where measurement coherence equals zero, which is equivalent to no feedforward (feedback-only) control. Results show the benefit of wind turbine preview control as a function of measurement coherence and preview time.