Data-driven real-time risk assessment of resilient distribution system during typhoon weather

Abstract

Coastal distribution systems are vulnerable to typhoon weather. An efficient and accurate real-time risk assessment is very necessary in providing timely warning information for decision making. However, the conventional risk assessment methods are subject to varied typology, parameters and uncertainties. While in current data-driven methods, the effect of complex factors such as resilient resources and emergency response are difficult to be analytically incorporated in risk assessment procedure. To solve the problems, this paper proposes an improved data-driven real-time risk assessment method for resilient distribution systems during typhoon weather. A comprehensive risk-oriented database with 25 indexes is constructed based on practical experience in system operation. And then the effects of various complex factors on distribution system risk are characterized by entropy weights and gray correlation degrees. The case study on a modified 33-node system has validated the proposed method in real-time risk assessment. The whole scheme can be helpful for the software and hardware update of resilient distribution systems.