An Intelligent Switching State Recognition Method Based on Multimodal Knowledge Fusion
Xunting Wang, Bo Gao, Bin Xu, Jinjin Ding · 2024
In order to solve the problems of low degree of automation, low identification efficiency and inability to identify safety risks in time, this paper proposes a switching state identification method based on multi-modal knowledge fusion. In this paper, a single shot multi box detector object detection algorithm using deep learning and linkage with SCADA system is introduced, which realizes high-precision state recognition of substation switchgear by combining lightweight visual image processing algorithm and deep neural network. By combining the remote signaling and telemetry information of the switching state obtained by the SCADA system, this method can quickly identify the switching state of the substation equipment, improve the accuracy and efficiency of the safety management and control of the power system equipment, ensure the reliability of the equipment status monitoring results, and maintain the safe and stable operation of the power grid.