Single Line-to-Ground Fault Special Protection Scheme for Integrated WindFarm Transmission Line Using Data Mining

Abstract

The need to solve the protection compromised currently preventing the smooth coexistence of both conventional fossil generation sources and the renewable green wind farm energy resources (WFER) on the same transmission line during a short circuit fault is the motivation for this study as the solution for meeting the pending future energy sustainability problems. The fast rate of global fossil-fuel reserves depletion, price instability, and climatic impact from the greenhouse gas (GHG) emission levels considered. A novel hybrid Wavelet-Machine Learning (W-ML) special protection scheme with the adoption of extracted 1-cycle wavelet decomposed transient fault signals features in Matlab/Simulink. The result from the supervised learning in the Waikato environment for knowledge analysis (WEKA) software indicated the best performance from Nearest-Neighbours (Lazy.IBK) classifier algorithm with 99.86 % classification for single-line-to-ground (SLG) faults, RMS error value of 0.0322 and instantaneous tripping time. The protection compromise is addressed for the effective future network coexistence.