Dual-Side State Identification Technology for UHV Converter Stations Based on Cyber-Physical Integration

Due to the wide range of data sources, high sampling density, diverse measurement devices, and complex communication protocols in ultra-high voltage (UHV) converter station systems, existing technologies struggle to accurately identify the complex states of converter stations and their underlying fault characteristics. In response to this challenge, this paper proposes a feature recognition technology for UHV converter stations based on cyber-physical integration. After preprocessing and performing correlation analysis on multisource heterogeneous data, mainly in the form of images, the method utilizes a cyber-physical dual-side state operation and transfer feature correlation matrix to analyze, train, and identify physical and communication-side faults within the converter station. A case study and method comparison were conducted based on actual monitoring images from a converter station. The results demonstrate that this method outperforms several traditional fault diagnosis and risk recognition algorithms, offering superior diagnostic capabilities.