Research on Infrared Intelligent Monitor Device for Transformer

Transformers play a central role in voltage regulation and electrical isolation. During operation, transformers are prone to overheating and other defects, which may result in electric power accident. Traditional infrared inspection methods for transformers rely on manual checks, it is possible to overlook some defects. Transformers possess complex structures and may develop various types of thermal defects after prolonged operation. Current monitoring research lacks the capability to conduct separate analyses for different types of defects. To address these limitations, this paper proposes an intelligent transformer monitoring device based on artificial intelligence and infrared thermography. This paper implements real-time monitoring device of substation transformers by employing the YOLOv8 object detection model integrated with defect detection algorithms on the NVIDIA Jetson Orin Nano platform. The proposed device enables lightweight operation of defect detection models while maintaining real-time performance and high accuracy. This capability facilitates early defect detection during transformer operation, thereby enhancing grid stability and operational intelligence.