Research on data-driven, multi-component distribution network attack planning methods

Abstract

As the physical power information system undergoes continual advancement, mobile energy storage has become a pivotal component in the planning and orchestration of multi-component distribution networks. Furthermore, the evolution and enhancement of big data technologies have significantly contributed to enhancing the rationality and efficacy of various distribution network planning and layout approaches. At the same time, multi-distribution networks have also confronted numerous network attacks with increasing probability and severity. In this study, a Petri net is initially employed as a modeling technique to delineate the network attack flow within the distribution network. Subsequently, the data from prior network attacks are consolidated and scrutinized to evaluate the vulnerability of the cyber-physical system (CPS), thereby identifying the most critical network attack pattern for a multi-component distribution network. Following this, the defender–attacker–defender planning methodology is applied for scale modeling, incorporating rapidly evolving mobile energy storage into the pre-layout, aiming to mitigate the detrimental impact of network attacks on the power grid. Ultimately, the column and constraint generation (C&CG) algorithm is utilized to simulate and validate the proposed planning strategy in a 33-node system with multiple control groups established to demonstrate the viability and merits of the proposed strategy.