A Hierarchical Data–Driven Wind Farm Power Optimization Approach Using Stochastic Projected Simplex Method

In a wind farm, the interactions among the wind turbines through wakes can significantly reduce the power output of the wind farm. These together with the complex wind conditions make the power optimization problem of the wind farm very challenging. To address this problem, this article proposes a hierarchical data–driven power optimization scheme, which does not need a wake interaction model that can be rather difficult to develop due to the complex aerodynamics between the turbines. The proposed scheme consists of two steps: firstly the power optimization problem of the wind farm is divided into several optimization sub-problems to deal with the complex wind conditions based on the wind farm power efficiencies in different wind directions. Secondly, a data–driven stochastic projected simplex algorithm is developed to solve the power optimization sub-problems. The proposed algorithm can increase the power output of the wind farm by using measurement data only and has the ability to find the optimal solutions. Finally, simulation results show that the proposed scheme can efficiently improve the power output of the wind farm in different wind conditions compared with some benchmark methods.