Integrated AGC Error Operation and Abnormal Instruction Online Identification

With the continuous advancement of power grid automation, current online identification of Automatic Generation Control (AGC) Error Operations and Abnormal Instructions presents several deficiencies, including inadequate coverage of identification methods, incomplete feature extraction, and insufficient in-depth analysis of online identification results. This paper establishes an integrated AGC online identification system to address these issues. First, rational/abnormal feature extraction is performed through multi-source information correlation analysis, employing logical language representation and knowledge graph representation. Second, the Significance of Data analysis method is utilized to determine the criticality levels of AGC input/output information. Finally, for various abnormal scenarios, the system integrates expert scoring, statistical analysis, and complex network methods to determine online identification probability and alert classification, ultimately generating operational outputs. Case analysis validate the system's effectiveness in simultaneously handling Error Operations and Abnormal Instructions. The proposed integrated online identification system for AGC Error Operations and Abnormal Instructions effectively mitigates existing deficiencies, significantly enhancing the reliability and operational stability of AGC systems.