Comparison of ANFIS and ANN Techniques in Fault Classification and Location in Long Transmission Lines

Abstract

This paper presents application of ANFIS and ANN in fault classification and location in a long transmission line. Compared to other methods, Machine Learning techniques based on artificial intelligence perform the best in fault classification and finding its location. Most frequently used ML techniques for this purpose are ANFIS and ANN. Both the techniques were able not only to identify fault type but also to find the fault location in the transmission line very accurately using source end current and voltage data. Common training and testing data was used for ANFIS and ANN. This data was obtained from simulation of faults in a long transmission line model using MATLAB. Error analysis and comparison of both the techniques is also presented in this paper. A GUI was designed for comparison of both the methods.