Research on Transformer Temperature Rise Detection Based on Faster RCNN

During the operation of power system, the thermal effect of electrical equipment is the precursor of many faults. In order to ensure the safe and stable operation of power equipment, it is very important to carry out timely temperature detection and fault analysis. As an important equipment in power system, it is very important to ensure the safe and stable operation of current transformer. In order to avoid the influence of current transformer fault heating, this paper establishes a current transformer detection and recognition mechanism, and proposes a current transformer recognition system based on image detection. It has strong spatial feature extraction ability, and the model simulation recognition degree reaches 100 %. The field experiment analysis shows that the system has good stability and strong accuracy. It can accurately detect and identify the current transformer and its temperature, which is of great significance to engineering practice.