Infrared Detection of Thermal Faults Inside Transformers Based on Higher Order Harmonic Detection and Independent Component Analysis

Jingwei Yan,Fuquan Jin,Changjin Chen,Chunyue Jiang,Weitao Hu,Haibo Guo
DOI: https://doi.org/10.1117/12.3004289
2023-01-01
Abstract:The current technology for detecting thermal faults within transformers is relatively simple in its analysis of infrared thermal images, resulting in poorly correct detection results. To this end, an infrared detection technique for thermal faults inside the transformer based on higher order harmonic detection and independent component analysis is proposed. The thermal images of the internal transformer are acquired. Extraction of higher order harmonic distortion signal waveforms from thermal images. Transmission of the distortion signal waveform to obtain the minimum set of attributes for independent component analysis. Corresponding to the fault site, the detection of thermal faults inside the transformer is completed. Experiments have shown that the technique has a high correct rate of 99.57% on average for different types of thermal faults inside the transformer, which is of high application value.
What problem does this paper attempt to address?