Resumen de Control Inteligente del ángulo de palas de un aerogenerador mediante aprendizaje por refuerzo

Jesús Enrique Sierra García, Matilde Santos Peñas, Inazio Cavero

• In this work, reinforcement learning has been usedto train a wind turbine pitch controller. The selected algorithmhas been DDPG, and it was implemented in a simulation environment.The controller generates collective pitch angle references,to obtain a smooth rotor speed.Preliminary results are presented, which demonstrate the potentialof this approach to solve the pitch control problem.