Online Status Assessment o Distribution Network Transformer Based o Random Matrix Theory

Aiming at insufficient representation of current indicators in transformer health assessment, this paper, using random matrix theory (RMT), motivates data-driven tools to analyze and perceive the spatiotemporal correlation of distribution network transformer equipment operation data, and then constructs high-dimensional statistical characteristics as the representation quantity of equipment health status. Some actual cases are given to show that the proposed feature has a better characterization ability for the equipment health condition with consideration of sensitivity and reliability. The work through the object – data – high-dimensional statistics – health index, on the one hand, to achieve data-driven health assessment, on the other hand, to provide theoretical support and application demonstration for the implementation of big data analytics in the field of power systems.