A Protection Scheme for Fault Identification in AC Microgrid Using LDA-Based Algorithm

Due to the variety of fault scenarios in microgrids, fault identification is an essential requirement to quickly minimize any major systemic disruption. In microgrids varying fault scenarios and the occurrence of dissimilar AC faults can rapidly change the fault current level from low to high which can affect the microgrid performance. The protection scheme based on linear discriminant analysis is presented in this research to identify the fault in microgrids under the above scenarios. The proposed microgrid protection scheme has been validated for mode identification, detection and classification of fault along with section identification under diverse operating conditions. The voltage and current samples have been taken from the selected bus for processing data using discrete wavelet transform under both the operating modes, i.e., grid-connected and islanded modes. Some cases of weather-based random scenarios have also been considered for validation. Results indicate about robustness and accuracy of the protection scheme.