Optimal dispatch of reactive sources in wind farms

This paper suggests three different approaches for online reactive power dispatch in wind farms by using a new heuristic optimization algorithm called Mean-Variance Mapping Optimization (MVMO). Optimization for any given operating point will result in minimum losses but doesn´t allow the consideration of the cost of on-load tap changer (OLTC) movements. To solve this problem the authors propose a predictive optimization where the objective is extended by the number of OLTC tap changes. Besides several time steps ahead are optimized simultaneously by using the information provided by a short term power forecast. In addition the authors suggest a method for incorporating the reactive power optimization into the wind farm control loop so that it can be used online to determine the optimal distribution of reactive power generation by each wind turbine. The proposed methods are demonstrated in a representative offshore wind farm.