A Method for Quickly Identifying Abnormal Status of GIS Equipment in Digital Twin Scenarios

Digital twin technology, with its characteristics of virtual-real integration, is becoming a key means to improve production efficiency, ensure stable operation of equipment, and realize fault warning. This paper proposes a method for rapid identification of abnormal states of GIS equipment in digital twin scenarios. First, the abnormal characteristics of online data related to the gas chamber of the GIS switch unit are analyzed, and the fluctuation coefficient that characterizes the dispersion of state quantity data is inferred by using a sliding window combined with the Weibull distribution model. Weights are assigned to each characteristic state quantity according to the fluctuation coefficient. The state classification of the state quantity is realized according to the weighted Euclidean distance combined with the optimized Canopy algorithm. The abnormal event library is established to realize the abnormal state identification of the gas chamber of the GIS switch unit, which can realize the rapid identification of abnormal states of GIS equipment in digital twin scenarios.