Machine Learning Based Intentional Islanding Algorithm for DERs in Disaster Management

Currently, research work is primarily dependent on the collection of large sets data from systems and making predictions based knowledge obtained data, which generally termed as `data mining'. These mining algorithms are great importance in improving performance different applications. In this regard, Machine Learning (ML) have been demonstrated to be excellent tools cope with difficult problems. paper, a classification learner supervised ML algorithm proposed for intentional islanding DERs live collected supervisory control acquisition (SCADA) system post disaster situations. Literature presents various detection techniques also address problems AC networks. majorly current source or voltage inverters. On other hand, low DC distribution allowing removal inverter proposed, supposed more advantageous by reducing losses economical when working DERs. considering effects natural disasters. The models trained fine tree, linear SVM, quadratic SVM Gaussian SVM. training tree model achieved higher accuracy 99.8%. main objective achieve faster accurate decision making. compared earlier artificial intelligence (AI) algorithms. AI fuzzy inference (FIS), neural networks (ANN) adaptive network (ANFIS). comparison shows that, than all