Detection and classification of multiple power quality disturbances in Microgrid network using probabilistic based intelligent classifier

Microgrid (MG) networks have evolved as reliable power source for providing secure, reliable, and low carbon emission of energy supply to the remote communities. Power quality disturbance (PQD) is a common issue affecting the performance of the MG network and hindering its usage in small scale. PQD tends to lessen the reliability, performance, and life-cycle of the various power devices in the network. Hence, in this study, a probabilistic based intelligence method has been proposed to detect and classify the PQDs more accurately in the MG network. MG system has been developed using built in features available in the Matlab/Simulink platform. Discrete Wavelet Transform (DWT) based signal processing technique has been applied to extract the features from the multiple PQD signals. The obtained features are used to train the computational intelligent based classifiers such as Multi-Layer Perceptron (MLP) neural network, Support Vector Machine (SVM), and Naive Bayes (NB). The results obtained indicate the proffered NB and SVM classifier could predict PQDs in the MG network with 100% classification accuracy while the MLP gives the classification accuracy of 66.7%. Further, the robustness of classifiers is evaluated using performance indices (PI) of Kappa statistic, mean absolute error and root mean square error. From the PI evaluation, it can be concluded that the probabilistic based NB approach gives the predominated result compared to SVM and MLP method.