Presentation Description: Wind farms today consist of an array of independently controlled turbines. At Arctura, we envision a future where wind turbines are collectively controlled to optimize yaw angles with the objective of continuously maximizing the total power generation at below-rated wind speeds. Known as “wake steering,” the approach can increase the annual energy production of existing wind farms by 1-3% with marginal cost increase, thereby reducing LCOE and increasing profit for the owners and operators. Prior to new construction, confidence in a wake steering approach could allow tighter turbine spacing, potentially reducing the balance-of-plant capital costs for offshore wind farms. While some research groups are pursuing wake steering using AI and machine learning algorithms, or model-based supervisory control schemes, Arctura is developing a novel method that we believe will prove to be more robust, easier to implement, and less costly than these other approaches. Our method uses a model-free control framework known as Extremum Seeking Control (ESC), which is a popular control method used in automotive and other industries. The unique algorithm being developed by Arctura overcomes certain limitations of the more general ESC algorithm which make it difficult to apply on wind farms in variable wind conditions. In this presentation, we will review the genesis of the concept, the development and validation process, and we will review our planned wind farm field tests.
Learning Objectives:
Upon completion, participant will be able to describe the impact of turbine wakes on array performance.
Upon completion, participant will be able to describe the underlying Extremum Seeking Control framework in general terms.
Upon completion, participant will be able to describe the innovation required to make Extremum Seeking Control work for wind farms.
