Chapter 31 - Identification and classification of faults using Stockwell transform and decision rule

Abstract

The electrical power grids are becoming large in nature and are being transformed into the smart grid by using and deploying the technology of smart grids. This raised the requirement of improved/advanced schemes for protection of these complex grids, which require fast and accurate estimation of the faults. Fault estimation included the identification and classification of the various faults. This also includes the discrimination of phases with faulty from the healthy phases. The research activity included in thesis is aimed on the designing an algorithm based on the processing of current signals using Stockwell transform for identification of faults on the power transmission line. The classification of the faults has been achieved using the rule-based decision tree. Current signals recorded during the faulty condition are processed using the Stockwell transform to obtain output matrix. This matrix is used to compute the indices such as median intermediate fault index (MIFI), maximum value intermediate fault index (MVIFI), summation intermediate fault index (SIFI), and covariance fault index (CFI). A hybrid fault index (HFI) is computed by multiplying the MIFI, MVIFI, SIFI, and CFI. A threshold value for the hybrid fault index (HFIT) equal to 200 is selected for the proposed HFI to estimate the faulty condition from the healthy conditions. This HFIT is decided by the testing the algorithm for the different conditions such as different values of the fault location, fault impedance, fault incidence angle, reverse power flow, etc. The faults have been effectively classified using the rule-based decision tree with the help of CFI.