A Visualization Verification Method for Startup Risk of Substation Equipment Based on Multidimensional Graph Models

作者：Lifan Mao, Huang Chen, Yongqi He, Chuanpu Zhu, Chao Chen · 发表于：Engineering Reports · 年份：2025 · DOI：10.1002/eng2.70471 · 被引用次数：1 · 研究领域：Power System Reliability and Maintenance、Risk and Safety Analysis、Power Transformer Diagnostics and Insulation

ABSTRACT This article presents a visualization verification method for substation equipment startup risk based on a multidimensional graph model, aiming to enhance the safety and reliability of the equipment startup process. By integrating the common information model of substation equipment with scalable vector graphics files, a PMS 2.0 model file is constructed, and a multidimensional model database—comprising a model library, a front‐end library, and an offline library—is established. Upon inputting a real‐time image of the device to be verified, an adaptive registration method based on feature points is employed to align it with the reference image from the model library, after which a device topology model is constructed. Lightweight YOLOv5 algorithm is integrated to achieve intelligent verification and visual output of equipment startup risks. Experimental results demonstrate that the proposed method can effectively identify startup risks in key equipment such as transformers, offering an intuitive and reliable risk assessment reference for operation and maintenance personnel, and demonstrating high value for engineering applications.