Wind Power Generator Model Based on LS-SVM for Unbalanced Three-Phase Distribution System Power Flow Studies

Wind power generation is increasing around the world. However, increasing wind power penetration has a significant impact on the power system, in particular on distribution systems. Accurate wind generator models are needed for proper unbalanced load flow analysis. A non-linear equation-based model is the most accurate as it models, though requiring greater computational power and time. Machine learning-based models, such as Artificial Neural Networks, are accurate enough and are faster than non-linear models. This paper explores another machine learning algorithm-Least Squares Support Vector Machines (LS-SVM)-in wind generator modeling. It was shown that the LS-SVM model is more accurate than the artificial neural network (ANN) model and almost as accurate as the non-linear model. When the LS-SVM model was integrated to the load flow analysis of the IEEE 37-bus system, the solution had little to no deviation from the results using the non-linear model of wind generators.