Intelligent Islanding Classification with MLPNN for Hybrid Distributed Energy Generations in Microgrid System

Due to the rise in penetration of distributed energy resources (DERs) and energy storage systems (ESSs) into the microgrid (µG) system, the abrupt disconnections of DERs and ESSs might have an effect on the stability and dependability of the entire µG system. It is important to take into account the low-voltage ride-through (LVRT) capabilities to keep the µG system operating steadily. In order to quickly identify grid abnormalities and categorize islanding scenarios, this paper attempts to develop an islanding detection technique. The procedure involves generation of data samples corresponding to different grid abnormalities and training them with a machine learning based classifier. To achieve this process, a grid-connected (GC) three-phase µG system is simulated and compiled in Typhoon HIL application software to get distinct data sets for different grid fault conditions. Additionally, the data is trained using the multilayered perceptron neural network (MLPNN) classifier to create the islanding classification process. The trained classifier has 99.5% effectiveness and can classify data at a rate of about 170000 observations per second.