A comparative analysis of different mother wavelets for fault detection and classification in the Nigerian 330 kV transmission network

Transmission line fault identification and classification are necessary to ensure the stable and reliable operation of the power systems. During a fault, power flow is disrupted, resulting in a transient condition. Tripping action on the transmission line is mainly dependent on current and voltage waveforms, which are sometimes obtained at the relay location. In order to identify and classify these transmission line faults in real time, fast and accurate analysis is required. Many researchers have deployed signal processing algorithms as tools to study the voltage and current waveforms of fault signals. This paper focuses on a discrete wavelet-based technique for fault detection and classification using the Nigerian 330 kV transmission network as a case study. Detailed coefficients of three-phase and ground currents were used to capture high-frequency transients generated at the instant of fault occurrence (0.4 s). A threshold value of 40 was selected, compared to 31.551 maximum coefficient at no-fault. Ten mother wavelets were comparatively evaluated using the maximum detail coefficient as the performance index. Simulation results show that while all wavelets successfully detected and classified faults, higher-order wavelets such as db4, db8, sym4, and sym8 provided improved stability, lower normal-condition baseline values, and superior fault discrimination. MATLAB/SIMULINK software is used to test the model system with the proposed approach, demonstrating fast response, computational simplicity, and suitability for real-time transmission line protection.