Model predictive control attenuates wind shear- and tower blockage loads in a wind turbine drive-train
Year: 2026
Type: conference-paper
Abstract: Constantly-fluctuating loads cause fatigue damage in wind turbine structural components, leading to costs associated with maintenance and replacement. The purpose of this study is to develop control algorithms that reduce vibrations in wind turbine drive-trains. Results of this work show that the tower’s blockage of the wind, and to a lesser extent, vertical wind shear, are important contributors to drive-train loads, that they can be predicted with a simple model using only the generator speed-, blade pitch-, and rotor azimuth signals, and that predictively controlling for them reduces loads 7% more than simply reacting to them with a feedback controller. For the 5MW NREL Offshore Baseline turbine controlled by the ROSCO controller, model-predictive drive-train damping achieves 37% reduction in low-speed shaft damage-equivalent load, vs. 32% for linear-quadratic Gaussian feedback. Control-related computations use 0.02% of the simulated 62.5ms sample period for both feedback- and predictive controllers; for comparison, the ROSCO controller uses 0.05% of its 12.5ms sample period. The predictive control design method developed here shows potential for extension to reduce blade and tower loads by compensating for rotor mass-and aerodynamic imbalances. Additionally, this paper introduces an open-source, cross-platform framework for the development of- and communication between control- and signal-processing algorithms, with interfaces to many wind turbine aeroelastic simulation programs (https://github.com/jaehlers/dispatch).
Author: Julian Ehlers
Topic: Wind Turbine Control Systems
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7162692541