Distribution grid fault analysis under load and renewable energy uncertainties

A non-transitory computer readable medium having instructions stored therein that, when executed by one or more processors, causes the one or more processors to perform a method for intelligent fault diagnosis of a distribution feeder connected to renewable energy resources, comprising:

• modelling, with a computing system having circuitry configured for modelling and processing, the distribution feeder, the distribution feeder having predetermined electrical characteristics, wherein the distribution feeder is divided into branches separated by nodes, wherein each node is connected to at least one of a power input or a power output, wherein at least one power input is a renewable energy resource and at least one power input is a main feeder;
• modelling uncertainties in the energy supply of at least one renewable energy resource connected to a power input node by a first probability density function describing a load demand of at least one power output node and a second probability density function modelling uncertainties in the energy supply of the at least one renewable energy resource;
• generating, from the modelling, a first dataset of three phase current signals for a plurality of pre-specified branches of the distribution grid;
• generating, from the modelling, a second dataset of three phase current signals of the main feeder and at least one renewable energy resource;
• analyzing the first dataset and the second dataset to extract features of each branch;
• recording the current signals at each power output node;
• determining whether a branch has a fault by comparing the current signals at each power output node to the extracted features of each branch;
• locating and classifying the fault;
• and
• displaying, by a display, the features of the fault and the fault location.