Enhanced Fault Type Detection in Covered Conductors Using a Stacked Ensemble and Novel Algorithm Combination.

Area of Science:

• Electrical Engineering
• Power Systems

Background:

• Cross-linked polyethylene (XLPE)-covered conductors are crucial for reliable power distribution, especially in forested areas, reducing buffer zones.
• Existing fault detection methods are often expensive, difficult to install, and lack detailed fault type information.

Purpose of the Study:

• To develop an innovative, antenna-based approach for precise fault detection in XLPE-covered conductors.
• To classify various fault types, including single-phase, three-phase, and different fault categories, which was not previously possible.
• To enhance the accuracy and reliability of fault detection systems for power distribution networks.

Main Methods:

• Utilized an antenna-based system for fault signal detection.
• Employed a classifier to differentiate between various fault types.
• Implemented a stacking ensemble method with logistic regression to improve detection accuracy.

Main Results:

• Successfully differentiated between single-phase to three-phase faults.
• Achieved classification of different fault types, a novel capability.
• The stacking ensemble method significantly bolstered the accuracy of fault detection.

Conclusions:

• The proposed antenna-based method offers a significant advancement in fault detection for XLPE-covered conductors.
• This technology can lead to reduced false positives, improved network reliability, and safer utilization of covered conductors.
• The findings support wider adoption of covered conductors, contributing to environmental protection and efficient power distribution.