Fault Detection AND Classification in Transmission Lines using Boosted Decision Tree

Abstract

Fault detection and classification in transmission lines is a crucial task for engineers to maintain reliability and safe operation of electrical power systems. This article proposes a new technique based on statistical features and Boosted Decision Tree (BDT) to identify the fault and classify it. The essential statistical features are calculated from fault currents with 10 different types of faults, then BDT is applied to identify and classify the faults. Experimental results show that the proposed technique can identify and classify transmission line faults accurately. The proposed BDT is compared with other competitive machine learning classifiers to justify the improved performance.