Open AccessPrediction of overcurrent rRelay miscoordination time using artificial neural network
Author(s) -
S. Karupiah,
M. H. Hussain,
Ismail Musirin,
Siti Rafidah Abdul Rahim
Publication year - 2019
Publication title -
indonesian journal of electrical engineering and computer science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.241
H-Index - 17
eISSN - 2502-4760
pISSN - 2502-4752
DOI - 10.11591/ijeecs.v14.i1.pp319-326
Subject(s) - relay , overcurrent , artificial neural network , computer science , matlab , protective relay , multiplier (economics) , digital protective relay , power (physics) , engineering , artificial intelligence , electrical engineering , voltage , physics , quantum mechanics , economics , macroeconomics , operating system
Overcurrent relay plays an important role in the protection of power system. For protection, proper coordination of relays with an appropriate relay settings need to be done. Coordination can be done by selecting an optimal Time Multiplier Setting (TMS) and Plug Setting (PS) considering the fault current at the relay location. Continuous Time Intervals (CTI) must be maintained between primary relay and secondary relay to ensure correct sequential operation of the relays. However, miscoordination can occurs due to secondary relay trips faster than primary relay. This paper presents an approach for predicting overcurrent relay miscoordination time using Artificial Neural Network (ANN) algorithm in MATLAB software. The efficiency of the proposed approach have been tested successfully on 17 bus test system. The simulation results indicated that the ANN Levenber-Maequardt algorithm is capable to predict the miscoordination time occurs between the primary and secondary relay operating time.