An improved fault detection classification and location scheme based on wavelet transform and artificial neural...

Area of Science:

• Electrical Engineering
• Power Systems Engineering

Background:

• Six-phase transmission systems offer increased power transfer capacity with minimal infrastructure changes, addressing right-of-way restrictions and rising energy demands.
• The widespread adoption of six-phase systems is hindered by the lack of robust protection schemes due to the complexity of fault analysis.

Purpose of the Study:

• To develop and validate a novel, reliable, and efficient fault detection, classification, and location scheme for six-phase transmission lines.
• To utilize single-ended data for protection, simplifying installation and data acquisition.

Main Methods:

• A hybrid approach combining discrete wavelet transform (DWT) for signal decomposition and modular artificial neural networks (ANN) for fault analysis.
• Utilizing the standard deviation of approximate coefficients from voltage and current signals as input features for the ANN.

Main Results:

• The proposed scheme successfully identified and located all 120 types of shunt faults under various conditions, including different fault locations, resistances, and inception angles.
• The scheme demonstrated robustness against variations in power system parameters such as source short-circuit capacity, X/R ratio, voltage, frequency, and CT saturation.

Conclusions:

• The hybrid wavelet transform and modular artificial neural network-based protection scheme is effective and reliable for six-phase transmission lines.
• The proposed method is suitable for real-time implementation, paving the way for increased acceptance and deployment of six-phase transmission technology.