Fault Detection and Classification for Compensated Transmission Lines with UPFC and Wind Farm Using Artificial Intelligence

Resumen

The world is moving towards renewable energy source to be the major source to meet the increasing electricity demand and wind energy is one of the suitable sustainable energy sources. This brings the need to integrate large-scale wind farms with the existing power grids. Doubly fed induction generator (DFIG) are most preferred wind turbine generators with benefits like high efficiency and flexible power. However, DFIG is sensitive to grid disturbances and so traditional relaying protection techniques are not applicable. DFIG are integrated to the power grid with transmission line. Using UPFC compensated transmission lines increases the complexity further. The grid side current signals are extracted using Variable Mode Decomposition (VMD) method. These results in grayscale images which are then are used in convolutional neural network (CNN) classifier for fault classification. The simulation is done using 2021 version MATLAB simulink with 1300 fault data. There are a total of 10 cases of fault. The performance of the deployed CNN classifier is assessed with accuracy being 97.95 % for CNN.