A reinforcement‐learning approach for individual pitch control

Individual pitch control has shown great capability of alleviating the oscillating loads experienced by wind turbine blades due to shear, atmospheric turbulence, yaw misalignment, or wake impingement. This work presents a novel controller structure that relies on separation low-level tasks and high-level ones. It is based neural network modulates basic periodic angle signals. trained with reinforcement learning, trial error way acquiring skills, in low-fidelity environment exempt from turbulence. The further deployed large eddy simulations assess its performances turbulent waked flows. Results show method enables learn how reduce fatigue exploit knowledge complex When compared state-of-the-art individual controller, one introduced here similar load alleviation capacities at reasonable turbulence intensity levels, while displaying very smooth pitching commands nature.