Arc Fault Detection for Photovoltaic Systems Using Independent Component Analysis Technique and Dynamic Time-Warping...

Area of Science:

• Electrical Engineering
• Renewable Energy Systems
• Signal Processing

Background:

• Arc faults in photovoltaic (PV) systems pose significant fire and explosion risks.
• Existing detection methods often rely on local arc features, leading to failures in noisy conditions.
• A global feature-based approach is needed for reliable PV arc fault detection.

Purpose of the Study:

• To propose a novel arc fault detection method for PV systems.
• To utilize global signal characteristics for enhanced detection accuracy and robustness.
• To address the limitations of traditional local feature-based methods.

Main Methods:

• Independent Component Analysis (ICA) to decompose DC current signals.
• Characterizing arc faults based on the independence of decomposed signals.
• Dynamic Time Warping (DTW) to quantify signal independence for detection.
• End-to-end arc fault detection system implementation.

Main Results:

• PV DC currents with arc faults decompose into two independent components via ICA.
• Normal PV DC currents do not exhibit this two-component independence.
• The proposed ICA-based method demonstrates superior detection accuracy.
• Enhanced robustness against environmental disturbances compared to traditional methods.

Conclusions:

• The proposed method effectively detects arc faults by analyzing signal independence.
• Global feature analysis using ICA offers a significant improvement over local feature methods.
• This approach enhances the safety and reliability of photovoltaic systems.