Feature Extraction of Internal Fault Types for the Solar PV System using the Novel NIFM Techniques

In low-voltage solar grids, traditional fault analysis and monitoring systems are used to detect power faults. These include wide-area protection technology with GPS and phasor measurement units (PMUs) and protective relays at each transmission line end. However, these designs increase data complexity and system cost. This paper uses the non-intrusive load identification technology, which involves installing a measuring device at the entrance of the extra high-voltage (EHV) transmission network. The measurement signal is processed by a proposed feature extraction method to form power energy features of various fault types in the frequency domain of the internal line of the low-voltage solar photovoltaic power grid. This approach is not only easier to install, but it is also more cost-effective. The results of the proposed method show the apparent differences for the different fault types.