The Study of Substation Equipment Fault Detection Using Infrared Imaging Technology
Zhuxing Ma, Lishuo Zhang and Jinkao Wang
The safe functioning of the power grid may be successfully ensured by accurate fault detection of substation equipment. This paper briefly introduced the infrared image-based substation equipment fault detection technology. Prior to using the relative temperature difference approach to assess the level of equipment failure, a classifier was first employed to identify the kind of equipment in the infrared images. Simulation experiments and field operation experiments were performed in the laboratory and in actual substations to test the recognition performance of the classifier and the diagnostic performance of the relative temperature difference method. The results showed that whether in simulation experiments or in field experiments, using a convolutional neural network as a classifier could more accurately identify the types of substation equipment in infrared images and using the relative temperature difference method could make more accurate diagnoses.
Keywords: infrared image, substation, fault diagnosis, relative temperature difference method