Intelligent fault diagnosis for distribution grid considering renewable energy intermittency

This paper proposes a versatile intelligent fault diagnosis (IFD) scheme for a distribution grid integrated with intermittent renewable energy resources (RER). Renewable generation parameters (wind speed and solar irradiation) and load demand intermittency along with fault information (fault inception angle and resistance) uncertainty are modeled by employing different probability density functions. Then, advanced signal processing techniques are used to extract useful features from the recorded signals. The proposed approach sends the extracted features as inputs to feedforward neural networks (FF-NNs) to diagnose (detect, classify, and identify faulty sections) and locate the faults. The presented results confirm the efficacy of the developed IFD scheme and show that it is independent of renewable generation and load demand intermittency along with fault information uncertainty. Additionally, the proposed scheme is independent of the presence of measurement noises. Furthermore, this work investigates the effectiveness of the developed IFD scheme under various contingency cases (branch outages and RER generation outages). Finally, a laboratory prototype IFD scheme is built by integrating a physical phasor measurement unit (PMU) with a real-time digital simulator (RTDS) rack to diagnose faults in the distribution grid. The results confirm the effectiveness of the prototype IFD scheme, as they show good agreement with the simulation results.