Fault detection on the transmission lines using fuzzy neural network

Duong Hoa An, Nguyen Thanh Thuy, Truong Tuan Anh

fault detection, time domain reflectometry, TSK neural network.

The faults can happen to transmission lines at any time, any places and caused by different reasons. An accurate and fast solution to detect, locate and isolate the faults will improve the quality of the power systems’ performance. The Time-Domain Reflectometry (TDR) method has been used to detect the fault location based on the time a pulse signal need to travel from the begin of the lines to the fault location and back. But due to the presence of non-resistance elements and nonlinear elements on the lines, the reflected impulse was deformed, which in turn reduces the accuracy of the travelling time calculation. In this paper, a Neuro-fuzzy network was used to improve the fault location detetion based on the analysis of reflected waveforms on the transmission lines after sending the pulse into the line. This paper presents the numerical results using Matlab-Simulink models to show the high quality of the proposed method.