Transmission line distance relaying using machine intelligence technique

Abstract

A new approach for distance relaying of transmission line using machine intelligence technique such as support vector machine (SVM) is presented. The proposed SVM technique is used for faulty phase selection and ground detection in different fault situations that occur on large power transmission line. Post-fault current and voltage samples for one-fourth cycle (five samples) are used as inputs to SVM 1, which provide output for faulty phase selection. SVM 2 is trained and tested with zero-sequence components of fundamental, third and fifth harmonic components of the post-fault current signal to provides the involvement of ground in the fault process. The polynomial and Gaussian kernel SVMs are designed to provide the most optimised boundary for classification. The total time taken for faulty phase selection and ground detection is 10 ms (half cycle) from the inception of fault. Also the proposed technique is tested on experimental set-up with different fault situations. The test results are compared with those of the radial basis function neural network and were found to be superior with respect to efficiency and speed. The classification test results from SVMs are accurate for simulation model and experimental set-up, and thus provide fast and robust protection scheme for distance relaying in transmission line.