Electric energy meter loop abnormal data identification method and apparatus, and electronic device

Abstract

The invention provides an electric energy meter loop abnormal data identification method and device and electronic equipment, and relates to the technical field of power systems. The method comprises the steps of obtaining time sequence electrical data of an electric energy meter loop; inputting the time sequence electrical data into an abnormal data identification model to obtain a data abnormal probability, the abnormal data identification model being obtained by constructing and training based on two layers of long and short-term memory networks and an attention layer; if the data anomaly probability indicates that the time sequence electrical data is abnormal data, a hidden state sequence and a gating signal in an abnormal data recognition model are obtained; and according to the data anomaly probability, the hidden state sequence and the gating signal, an abnormal data identification result is determined, and the abnormal data identification result comprises a time step importance sequence and a feature dimension contribution matrix corresponding to the time sequence electrical data, and a gating signal anomaly condition. According to the invention, the accuracy and the real-time performance of the abnormal data identification of the electric energy meter loop are improved.

Images