抽象的な

Fault Classification in Double-Circuit Transmission Lines Based on the Hierarchical Temporal Memory

BA Wokoma

In this paper a novel machine intelligence framework called the Hierarchical Temporal Memory is used for fault classification in double transmission lines. Fault location data estimation including associated transmission line parameter values are obtained via computer simulations. The fault location data generation problem is then reformulated into a multi-class state using a unique data transformation technique. The proposed technique is compared with two very popular state-of-the art machine learning algorithms – the Online Sequential Extreme Learning Machine (OS-ELM) and the Support Vector Machine (SVM). The results show that the proposed HTM model clearly outperformed the OS-ELM and SVM technique.