Identifikasi Gangguan Di Saluran Transmisi Tenaga Listrik Berbasis Metode Hibrid

Abstract

This research discusses a Hybrid Method for identifying short circuit faults in transmission line. The identification fault is done in the classification and estimation of fault location in Riau system (three buses). Hybrid Method is used such as Discrete Wavelet Transform (DWT) and Support Vector Machine (SVM). DWT is used to extract information of transient signals from each phase current and zero sequence current for one cycle when the fault begins. SVM classification is used to detect any fault in each phase and ground in determining the type of short circuit fault. SVM regression is used to estimate fault location that occurs in transmission line. The training and testing data are generated by simulating each type of short circuit fault in Simulink Matlab R2018a with variation of fault location, fault resistance and fault inception angle. Short circuit faults is applied in transmission line on simulink model from Koto Panjang (KP) bus to Garuda Sakti (GS) bus in Riau Province with voltage 150 kV and length 64 km. The obtained results are classification with accuracy 100 % and estimated fault location with the highest mean error value is 0.021529 %.
Keywords: Discrete wavelet transform, Estimate, Fault, Short circuit, Support vector machine.