Volume 25: Accelerated Energy Innovations and Emerging Technologies

Prediction Method of Dissolved Gas in Transformer Oil Based on Firefly Algorithm – Random Forest Xiu Zhou, Yunlong Ma, Yan Luo, Tian Tian, Weifeng Liu, Xiuguang Li, Ninghui He, Zhenghua Yan, Hui Ni



When a transformer fault occurs, the transformer oil will decompose and produce a large amount of dissolved gas in the oil, based on the dissolved gas in the oil to diagnose whether there is a fault in the transformer, known as dissolved gas analysis (DGA), in order to effectively predict whether a transformer fault will occur in the future, so as to prevent the development of the fault in time at the early stage of the fault, proposed A model for predicting the dissolved gas concentration in transformer oil based on the firefly algorithm (FA) optimized random forest (RF), which uses the random forest as the prediction model and adjusts the parameters in the RF by means of the firefly algorithm. The experimental results show that the FA algorithm can effectively optimize the parameters in the RF and improve the prediction accuracy of the model, overcoming the shortcomings of the traditional RF algorithm which uses random parameters with low accuracy, and the model can predict the dissolved gas concentration in oil more accurately than the existing methods.

Keywords Transformer, Prediction of dissolved gas in oil, Firefly algorithm, Random forest

Invalid URL for PDF Viewerhttps://www.energy-proceedings.org/wp-content/uploads/2022/08/MITAB_2022_paper_1182.pdf[/pdfviewer]

Copyright ©
Energy Proceedings