Wind Turbine Pitch Control First Approach Based on Reinforcement Learning
-
J. Enrique Sierra-García [1] ; Matilde Santos [2]
-
[1] Universidad de Burgos
Universidad de Burgos
Burgos, España
-
[2] Universidad Complutense de Madrid
Universidad Complutense de Madrid
Madrid, España
-
- Localización: Intelligent Data Engineering and Automated Learning – IDEAL 2020. 21st International Conference: Guimarães, Portugal; November 4–6, 2020. Proceedings / Cesar Analide (ed. lit.), Paulo Novais (ed. lit.), David Camacho Fernández (ed. lit.), Hujun Yin (ed. lit.), Vol. 2, 2020 (Part II), ISBN 978-3-030-62365-4, págs. 260-268
- Idioma: inglés
- Enlaces
-
Resumen
The control strategy defined for a wind turbine (WT) aims to achieve the highest energy efficiency and at the same time to ensure safe operation under all wind conditions. The goal of the pitch control of a WT is to stabilize the output power around its nominal (rated) value by means of the position of the rotor blades with respect to the wind. In this work, a pitch control strategy based on reinforcement learning (RL) is proposed. The control system consists of a state estimator, a reward mechanism, a policy table and policy update algorithm. Different reward strategies and policy update algorithms for the RL controller have been tested and compared with a PID regulator. The proposed controller stabilizes the output power of the wind turbine around the rated power more accurately and with smaller overshoot than the traditional one.
