Earth-Fault Detection Using Fuzzy Logic in Electrical Distribution Networks
Earth-Fault Detection Using Fuzzy Logic in Electrical Distribution Networks

Author(s): Razvan Moise
Subject(s): Energy and Environmental Studies
Published by: Udruženje ekonomista i menadžera Balkana
Keywords: Fuzzy logic; Expert systems; Earth-fault detection; Petersen coil; Admittance method
Summary/Abstract: The optimal functioning of electrical distribution networks is to maintain the voltage to acceptable limits in order not to affect the insulation of the electric lines. During a nominal regime the voltage variation is not so big due to the several voltage regulation methods, but during a fault regime, usually during earth-faults can occur overvoltages that affect the line insulation. That is why earth-fault must identified and eliminated as quickly as possible. In the present medium voltage networks, the earth-fault detection is made using the admittance and current injection method. That means that in the normal regime, the system measures the reference homopolar admittances of every electric line that is supplied from the substation busbar after each tunning cycle for the Petersen Coil. By injecting a current in the coil it can discover the line with the biggest asymmetry in a 20-30ms time interval, which is the earth-faulted line. This method is perfectly functional when is precisely known the homopolar current and the capacitive current of the grid. There are projects where the Holmgreen filter is chosen with a big nominal current and the measuring of the homopolar current is done with errors because of the small earth-fault current in the medium voltage networks with Petersen coil and this way the earth-fault detection can function with errors. The classic method needs precise data in order to have good results. The article is researching the possibility of detecting the earth-faults with fuzzy logic. The fuzzy logic uses artificial intelligence and allows earth-fault detection even if the input data is not so accurate because of the errors given by the instrument transformers. So using several logic rules, the phase and the circuit with earth fault can be successfully identified even if the input data is not so precise. The expert fuzzy systems can be successfully used in earth-fault detection and also in detecting any kind of fault with superior results to the classical method because using the data with errors can obtain good results using the logic small, normal, and big for currents and voltages in the system can take decisions similar to the human operator.

• Details
• Contents