A Novel Ultra-High Voltage Direct Current Line Fault Diagnosis Method Based on Principal Component Analysis and...

Area of Science:

• Electrical Engineering
• Power Systems
• Renewable Energy Integration

Background:

• Growing deployment of renewable energy necessitates robust protection for expanding direct current (DC) transmission grids.
• Traditional protection methods struggle with limited fault data and feature correlation, failing under remote or high-resistance faults.

Purpose of the Study:

• To develop an innovative single-ended protection principle for DC transmission lines.
• To overcome limitations of existing methods in detecting remote and high-resistance faults.

Main Methods:

• Employs Principal Component Analysis (PCA) for multidimensional feature extraction from fault data.
• Utilizes Kernel Density Estimation (KDE) to build a joint probability density function for fault identification.
• Integrates intrinsic correlations among multidimensional features for comprehensive fault diagnosis.

Main Results:

• The proposed PCA-KDE method demonstrates robustness against varying transition resistances and fault distances.
• Achieves 100% accuracy in fault identification across different sampling time windows (0.5 ms, 1 ms, 2 ms).
• Shows insensitivity to sampling frequency, enhancing practical applicability.

Conclusions:

• The novel single-ended protection principle effectively addresses limitations of conventional DC transmission line protection.
• PCA and KDE integration provides a superior approach for accurate and reliable fault diagnosis in modern power systems.