@article{article_1165004, title={FRT capability enhancement of wind turbine based on DFIG using machine learning}, journal={Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi}, volume={11}, pages={911–918}, year={2022}, DOI={10.28948/ngumuh.1165004}, author={Gencer, Altan}, keywords={Çift Beslemeli Asenkron Generatör (ÇBAG), Machine Learning (ML), Kapasitif Köprü Tipi Arıza Akım Sınırlayıcı (KKTAAS), Rüzgâr Türbini (RT)}, abstract={The doubly fed induction generator (DFIG) is very sensitive to the high voltage and current harmful effects that occur during the grid fault. A capacitive bridge type fault current limiter (CBFCL) based on the support vector machine (SVM), which is one of the machine learning (ML) methods, is presented to improve the fault ride-through (FRT) performance of in three phase-to-ground (3LG) symmetric grid fault that may occur in a wind turbine based on DFIG working under normal operating conditions in this study. The machine learning algorithm based on SVM has been implemented in both the control systems of DFIG converters and a control system of CBFCL. Four different SVM classifier algorithms are applied to generate the switching signals of electronic switching elements used in rotor side, grid side converter, and circuit topology of CBFCL. Fine Gaussian, Quadratic, Cubic and Linear kernel functions are preferred in the training of SVM classifiers. The developed SVMs have been suitably trained to true predict and decide behaviours of converters during normal and grid fault conditions. The performance of Fine Gaussian and Linear types of SVM is compared to the effectiveness of training efficiency for a wind turbine based on DFIG. The accuracy rate of the Fine Gaussian of SVM is 100 %, while the accuracy rate of Linear SVM is 22 %. The simulation results show that the Fine Gaussian SVM protects more efficiently from the harmful effects of 3LG grid fault compared to the Linear SVM for a wind turbine based on DFIG.}, number={4}, publisher={Niğde Ömer Halisdemir Üniversitesi}