Wind farm power output optimization using cooperative control methods

Abstract

AbstractWe study the application of cooperative control and game theoretic approaches to wind farm optimization. The conventional (greedy) wind farm control strategy seeks to individually maximize each turbine power. However, this strategy does not maximize the overall power production of wind farms due to the aerodynamic interactions (wake effect) between the turbines. We formulate the wind farm power optimization problem as an identical interest game which can also be used to solve other cooperative control problems. Two model‐free learning algorithms are developed to obtain the optimal axial induction factors of the turbines and maximize power production. The algorithms are simulated for a four‐turbine wind farm and the Princess Amalia wind farm and are compared to a learning algorithm that uses a game‐theoretic approach. It is shown that the proposed algorithms improve upon benchmark algorithm in terms of both performance and actuation effort.