Wind Energy Cultivating Vijnana

Introduction of self-learning yaw misalignment control intelligent entity, in MEE 2021.

Technical approaches to the wake challenge

caused by very complicated metocean conditions in the wind turbine for offshore

Wake caused by a wind farm decreases wind energy capturing efficiency, furthermore, in offshore, the wake with complicated metocean conditions reduces wind energy capturing efficiency much more. Via controlling yaw more precisely, the wake can be reduced, so the wind energy capturing efficiency can increase. There is a major factor of flow deflection among yaw controlling factors, and the flow deflection prediction model is non-stationary by the wake caused by a wind farm. Moreover, another major factor of assembly angle is also a non-stationary factor. The flow deflection prediction model’s non-stationary characteristics can be solved via self-learning based on deep reinforcement learning, and the assembly angle’s non-stationary characteristics can be solved via assembly angles estimated by differential Kalman filtering.

Refer to the PCT patent: Kalman Filter and Deep Reinforcement Learning based Wind Turbine Yaw Misalignment Control Method