AUTHOR=Sun Chenhao , Zhou Zhuoyu , Zhang Yongxi , Jia Zhiwei , Huang Jingjie , Huang Chenyang 

TITLE=A Dissolved Gas Assessment Model for Power Transformers According to Weighted Association Rule Mining

JOURNAL=Frontiers in Energy Research

VOLUME=Volume 10 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2022.879869

DOI=10.3389/fenrg.2022.879869

ISSN=2296-598X

ABSTRACT=As one indispensable part of power systems, the reliable-operated power transformers are vital for energy transmission, whereas are remarkably threatened by potential fault events. To achieve the satisfying and valid operation of power transformers, any fault events that may impact their health states ought to be evaluated and early warned. With such motivations, this paper presents original insights on the assessment of power transmission health states via their internal dissolved gas, and the association rule mining model incorporating the analysis of high-impact-low-probability components is proposed. In this model, the rarely occurred components in each feature can be assessed explicitly rather than being omitted. The principle of component importance measure is deployed to assess the corresponding weights of components via the risk from each input component rather than its data share. Finally, the parameters of the weight evaluation model can be dynamically adapted in an adjustment framework. This model is testified through an empirical case study, and the leading results can demonstrate its flexibility and robustness during real applications.