A DYNAMIC TRUST EVALUATION METHOD FOR BOUNDARIES OF NEW POWER SYSTEMS BASED ON MULTI-MODEL PARALLEL ANALYSIS
Keywords:
New power system, Network boundary, Dynamic trust evaluation, Multi-model parallel analysis, Risk identification, Grayscale controlAbstract
Aiming at the architectural characteristics of numerous nodes and fragmented access in new power systems, as well as the technical pain points of traditional boundary trust evaluation, such as difficulty in capturing short-term anomalies, identifying medium risks, perceiving complex threats, and poor model collaboration, a dynamic trust evaluation method for power system boundaries based on multi-model parallel analysis is proposed. This method constructs a four-level architecture consisting of data collection and preprocessing, multi-model parallel analysis, model collaborative fusion, and trust level output. It collects multi-dimensional boundary data through distributed probes and performs standardization processing, realizes full-complexity risk identification relying on three parallel layers of statistical, machine learning, and deep learning models, completes the fusion of multi-model results combined with a dynamic weight adjustment strategy, and finally maps to four-level grayscale trust levels with targeted grayscale control strategies. Experiments and applications show that this method controls the trust evaluation delay within seconds, realizing real-time, accurate, and grayscale evaluation of the trust status of boundary entities, and provides reliable technical support for boundary security protection of new power systems.References
[1] Sheng Gehao, Qian Yong, Luo Lingen, et al. Key Technologies and Application Prospects of Power Equipment Operation and Maintenance for New Power Systems. High Voltage Engineering, 2021, 47(9): 13.
[2] Cao Jinzhang, Zhu Chuanbai, Liu Bo, et al. Research and Implementation of Smart Grid Dispatching Coding System Based on Common Information Model. Automation of Electric Power Systems, 2011(2): 5.
[3] Han Ping, Zhang Han, Fang Cheng, et al. A Parallel Anomaly Detection Method for Massive Network Traffic Based on Multi-Model Fusion. Journal of Civil Aviation University of China, 2022, 40(1): 8.
[4] Zhu Yadong. Improvement of Boundary Node Identification Method in Cloud Computing Network. Computer Measurement & Control, 2017, 25(1): 167-169+172.
[5] Hu Xian, Feng Yiping, Pan Ge. Distributed LSTM Equipment Fault Prediction Based on Edge Side Autoencoder Compression//Proceedings of the 31st Chinese Process Control Conference (CPCC 2020), 2020.
[6] Chen Jianfei, Wang Rui, Zhang Fangzhe, et al. A Novel Anonymous Identity Authentication Scheme for New Power Systems Based on Ring Signature and IBE Strategy. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition), 2026, 38(01): 93-99.
[7] Feng Shijie, Qin Yanyan, Zeng Zhixiang, et al. Research on Trusted Operation and Maintenance for Power System Network Security Based on Zero Trust Technology. Network Security Technology & Application, 2025(09): 102-105.
[8] Xiong Kaizhi, Su Jianbo, Jiang Chao, et al. Identity Authentication Technology Based on Trusted Agent Under Zero Trust Architecture in Power Industry//Journal of Information Security Research. Proceedings of 2025 Cybersecurity Innovation and Development Conference. Yalong River Basin Hydropower Development Co., Ltd.; Huawei Technologies Co., Ltd., 2025: 124-127. DOI: 10.26914/c.cnkihy.2025.017943.