Identifikasi Gangguan Di Saluran Transmisi 115 KV Menggunakan Metode Twd Dan Recurrent Neural Networks
Abstract
The transmission system is the connecting part between the power plant and the distribution system and, will be forwarded to the load. When the fault occurred in the transmission line, it can cause a cut off in the supply of electrical energy to the load, causing losses to consumers. Therefore, a method is needed to identify fault accurately and quickly by reducing search time and speeding up the repair process. In this article, the identification of fault to be carried out is the classification and estimation of fault locations in the 115 kV transmission system. The method to be implemented is Discrete Wavelet Transformation (DWT) and Recurrent Neural Networks (RNN) using the type of Elman. This study applied the DWT and RNN methods to identify short circuit fault that occurs in the transmission line. DWT is aimed at extracting information of transient signals for each phase current and zero sequence current in one cycle when the fault occurred. RNN classification is used to detect a fault with each phase and ground and RNN to estimate the location of fault that occurs in the transmission line. Training and testing data generated by running simulations of each type of short circuit fault using Simulink Matlab R2016a with variation parameters; location point of fault, fault resistance and initial angle of fault. Short circuit fault applied in the transmission line on Simulink model of the Bus LK to Bus BK at a voltage of 115 kV and a line length of 63.3 km. The results obtained are a classification of fault with an accuracy of 100% and, estimation of fault locations with the highest average error value is 1.418568%.
Keywords: DWT, estimation of fault, RNN Elman, short circuit.
Keywords: DWT, estimation of fault, RNN Elman, short circuit.
Full Text:
PDFRefbacks
- There are currently no refbacks.