Abstract: 【 Purpose/significance】 Propose an innovative method to identify potential risk factors and potential accident escalation con⁃ sequences for the data stability of offshore oil and gas field power systems. Based on the principle of Directed Weighted Complex Net⁃ work (DWCN), integrate the Bowtie (BT) model, risk entropy, and improved Ant Colony Optimization (ACOSI) algorithm into this method. And determine the evolution of the accident from cause to effect, in order to identify the most likely path and quickly identify key data risk factors along the path. [Method/process] This article first qualitatively analyzes the evolution of data risk based on the BT model. Secondly, the evolution and development based on accident suppression can be divided into two stages. Finally, based on the ACOSI algorithm, the shortest path analysis of data risk evolution was conducted. Fuzzy set theory was introduced to calculate the fail⁃ ure probability of risk factors, and risk entropy was used to represent the uncertainty of risk propagation.【 Result/conclusion 】 This al⁃ gorithm is suitable for calculating the shortest path of data risk evolution and can be applied to specific situations in the stability of off⁃ shore oil and gas field power systems.【 Innovation/limitation 】The experimental results show that the data risk evolution method pro⁃ posed in this paper is an effective method for identifying the shortest evolution path and the most vulnerable risk factors, and can im⁃ prove the ability to identify data risks in the power system.