Smart fault detection and classification for distribution grid hybridizing ST and MLP-NN

Power quality disturbances become a major issue in modern commercial distribution grids, hence an innovative attempt to diagnose the faults is necessary for optimal management of power distribution grids and associated assets. This paper presents a hybrid approach using Stockwell transform (ST) and multilayer perceptron neural network (MLP-NN) to detect, and classify the faults in a simulated IEEE 13-node test distribution feeder in Real Time Digital Simulator (RTDS). In the proposed technique, the three-phase current waveforms are measured from different points in the feeder and then processed using ST to extract useful statistical features. The features are later fed into the MLP-NN system to detect and classify the faults. The approach proved to be highly efficient in terms of accuracy under both noisy and non-noisy measurements. In addition, the proposed approach is independent of pre-fault operating conditions as well as fault resistance and inception angle.