Abnormal Behavior Recognition Model of Power Grid Based on Multi-Scale Feature Fusion

Abstract

Abstract Artificial intelligence technology is applied to power grid transmission and operation status identification to improve the intelligence level of power inspection. However, the existing grid anomalous behavior recognition model has the problem of unsatisfactory accuracy for small target detection. To address this problem, this paper proposes an abnormal behavior recognition model of power grid based on multi-scale feature fusion. First, the texture feature extraction network with a multi-scale mechanism is constructed to obtain object features. Secondly, the graph relation network between target and background is constructed. Finally, the grid abnormal behavior recognition model based on multi-scale feature fusion is constructed. The model proposed in this article is compared with the existing object recognition model in the simulation of the different datasets. In terms of the evaluation accuracy index, the proposed model is 93.27, which is improved by 2.18%.