| Title |
A Study for the Improvement of the Fault Decision Capability of FRTU using Discrete Wavelet Transform and Neural Network |
| Authors |
홍대승(Hong, Dae-Seung) ; 고윤석(Ko, Yoon-Seok) ; 강태구(Kang, Tae-Ku) ; 박학열(Park, Hak-Yeol) ; 임화영(Yim, Hwa-Young) |
| Keywords |
FRTU(Feeder Remote Terminal Unit) ; Discrete Wavelet Transform ; Neural network ; Fault Indicator ; Inrush current |
| Abstract |
This paper proposes the improved fault decision algorithm using DWT(Discrete Wavelet Transform) and ANNs for the FRTU(Feeder Remote Terminal Unit) on the feeder in the power distribution system. Generally, the FRTU has the fault decision scheme detecting the phase fault, the ground fault. Especially FRTU has the function for 2000ms. This function doesn't operate FI(Fault Indicator) for the Inrush current generated in switching time. But it has a defect making it impossible for the FI to be operated from the real fault current in inrush restraint time. In such a case, we can not find the fault zone from FI information. Accordingly, the improved fault recognition algorithm is needed to solve this problem. The DWT analysis gives the frequency and time-scale information. The neural network system as a fault recognition was trained to distinguish the inrush current from the fault status by a gradient descent method. In this paper, fault recognition algorithm is improved by using voltage monitoring system, DWT and neural network. All of the data were measured in actual 22.9kV power distribution system. |