Detection and classification of open conductor faults in six-phase transmission system using k-nearest neighbour algorithm

The considerable rise in the demand of electrical energy in the existing power framework call for alternative solution. In this regard, six-phase transmission system seems to be the possible alternative which enables 73% more power handling capacity with existing transmission network. However, inspite of the associated advantages, protection of six-phase transmission line is still persisting as a great challenge. In this paper, a K-nearest neighbour (KNN) based protection scheme has been presented for detection and classification of all open conductor faults for a 138 KV six phase, 68 km long line simulated in MATLAB/Simulink environment. The fundamental components of current signals retrieved at the relaying point are used as input attributes to be fed to train the KNN. The effectiveness of the proposed scheme has been examined by variation in fault parameters such as inception angle and fault location. The observed results clearly reveal the performance of proposed technique in undertaking the intended task of detection/classification of open conductor faults in less than one cycle time. This validates the appropriateness and adaptability of the scheme under widely varying operating conditions.