2022 IEEE International Conference on High Voltage Engineering and Application (ICHVE 2022)

The aging state of the oil-paper insulation system inside the transformer is an important factor affecting the safe operation of the transformer. Considering that oil-paper insulation samples are difficult to obtain in large quantities, this paper first prepares oil-paper insulation samples by accelerated thermal aging test and then tested by Raman spectroscopy. Elman neural network is then used to simulate the real spectrum of the sample to obtain a large number of simulated spectra. Finally, all the Raman spectra obtained were used to build a diagnostic model for the aging of oil-paper insulation. The results show that the obtained simulation spectra are valid and the oil-paper insulation aging diagnosis model established in this paper can effectively evaluate the aging state of transformers. This paper develops a new way to solve small samples in oil-paper insulation diagnostic model, and at the same time lays the foundation for further quantitative assessment of the aging state of transformers.

oil-paper insulation,Raman spectroscopy,aging diagnosis,extended data,Elman nerual network,SVM (support vector machine)