In the power system, whether the transformer can work normally and stably will directly affect the safe operation of the power grid. Monitoring the real-time operational status of transformers is crucial for the early detection, diagnosis, and resolution of potential faults. In this paper, a fault detection method of oil-immersed transformer based on thermal imaging technology is proposed. Firstly, thermal imaging images of transformer under different working conditions are obtained by infrared thermal imaging technology. Then the feature extraction and fault detection of transformer thermal image are carried out by convolutional neural network. By conducting tests and validation on actual transformers, the accuracy of fault diagnosis has reached 98.7%, thus confirming the effectiveness and precision of this method.
本站僅提供期刊文獻檢索。
