Fault Type Detection and Classification Using Rule-Based Approach

Fault detection in power systems is critical for ensuring system reliability and stability. This study presents a rule-based classification approach for identifying fault types, including Single Line to Ground (SLGF), Double Line to Ground (DLGF), Line to Line (LLF), and Three-Phase to Ground (LLLGF) on IEEE 9-bus and IEEE 14-bus transmission systems. Voltage sag and current swell characteristics were extracted through simulations in PSCAD under varying fault resistances (0.01Ω to 70Ω) and fault distances. These features were analyzed in MATLAB using predefined logic rules based on amplitude thresholds and phase relationships. The classification method achieved 100 % accuracy for all fault scenarios on the IEEE 9-bus system. On the IEEE 14-bus system, the method maintained 100 % accuracy up to 50Ω fault resistance, but dropped to 73.75 % at 70Ω, particularly due to LLF cases with low fault current and voltage patterns resembling normal conditions. Overall, the classification algorithm reached $\mathbf{9 6. 2 \%}$ accuracy. The results indicate that the proposed method is effective for simpler topologies but faces challenges in more complex networks under high-resistance fault conditions. Future improvements may include signal processing enhancements or adaptive learning techniques to improve performance under these scenarios.