Autonomous Vision-Based Primary Distribution Systems Porcelain Insulators Inspection Using UAVs

Area of Science:

• Electrical Engineering
• Computer Vision
• Robotics

Background:

• Early detection of damaged outdoor insulators is vital for maintaining continuous electricity supply and public safety.
• Unmanned aerial vehicles (UAVs) offer a safer and more efficient alternative to traditional inspection methods, avoiding power system shutdowns.

Purpose of the Study:

• To develop and evaluate an advanced UAV-based system for the early detection of damaged outdoor insulators.
• To overcome data limitations by creating a novel dataset and employing image enhancement techniques.

Main Methods:

• A novel dataset was created using UAV-captured and manually generated images.
• Deep Laplacian pyramid super-resolution network and low-light image enhancement were used for image processing.
• A specialized fine-tuning strategy for object detection models was implemented, alongside optimized flight path strategies.

Main Results:

• The proposed method achieved high detection accuracy, with F1-scores of 0.81 and 0.77 on two distinct datasets.
• The system demonstrated robust fault recognition and diagnostic capabilities on real-world aerial inspection data.
• Proposed flight path strategies proved efficient in capturing video streams and minimizing the shuttering effect.

Conclusions:

• The developed UAV inspection system provides a simple, efficient, and accurate solution for detecting damaged outdoor insulators.
• The approach is validated through extensive evaluation on in-service porcelain insulators, showcasing its practical applicability.
• This method enhances power system maintenance by enabling early fault detection and diagnosis.