Active Wake Steering Control Data-Driven Design for a Wind Farm Benchmark

Upstream wind turbines yaw to divert their wakes away from downstream turbines, increasing the power produced. Nevertheless, the majority of wake steering techniques rely on offline lookup tables that translate a set of parameters, including wind speed and direction, to yaw angles for each turbine in a farm. These charts assume that every turbine is working well, however they may not be very accurate if one or more turbines are not producing their rated power due to low wind speed, malfunctions, scheduled maintenance, or emergency maintenance. This study provides an intelligent wake steering technique that, when calculating yaw angles, responds to the actual operating conditions of the turbine. A neural network is trained online to calculate yaw angles from operating data, including turbine status, using a hybrid model and learning-based active control method. The proposed control solution does not need to solve optimization problems for each combination of the turbines' non-optimal working conditions in a farm, as opposed to purely model-based approaches that use lookup tables provided by the wind turbine manufacturer or generated offline. Instead, the integration of learning strategy into the control design enables the creation of an active control scheme. The suggested methodology differs from solely learning-based approaches in that it doesn't call for a significant number of training samples, such as in model-free reinforcement learning. In actuality, by taking use of the model during backpropagation, the suggested approach learns more from each sample. Using a well-known and practical wind farm benchmark, results are reported for both standard (nominal) wake steering under operational conditions with all turbines and for faulty conditions.