Intelligent identification and early warning method for abnormal events in open scene in power field based on edge calculation

Abstract

An intelligent identification and early warning method for abnormal events in an open scene in the power field based on edge calculation compresses and transplants an improved SSD target detection model to a mobile terminal, gives full play to the advantages of the edge calculation, and takes an Android terminal as an optimal scheme through experiments. According to the method, the Conv4 _ x feature layer and the Conv5 _ x feature layer in the VGG16 network are fused, and then the fused feature layer directly acts on the final prediction layer, so that the accuracy of small target detection isimproved; meanwhile, various basic weather conditions such as sunny days, cloudy days, rainy days and foggy days are summarized, and the image enhancement technology is used for increasing training data under different scenes, so that the generalization ability of the model is improved.

Description

technical field [0001] The invention discloses an intelligent identification and early warning method for abnormal events in an open scene in the electric power field based on edge computing, and belongs to the technical field of intelligent identification of electric abnormal events. Background technique [0002] With the continuous development and construction of power engineering, the scale of the transmission network is getting larger and more equipment, and the inspection workload for transmission equipment and lines is also increasing. In addition, in some relatively remote areas, such as mountainous areas and snow fields, it is difficult for relevant staff to reach, which further increases the difficulty of inspection work. At the same time, the level of power grid management continues to improve, and the operation and maintenance units are also constantly exploring the operation and management methods of transmission lines. At present, the mainstream solutions for t...

AI Technical Summary

Problems solved by technology

[0005] Chinese patent document CN109165575A aims at the problems of high-speed rail surveillance video with complex background, high noise, and easy exposure of the picture, and studies a firework recognition algorithm based on image deep learning SSD framework, in which the detection model training network is the reconstructed VGG16 network. The detection model training network adds 6 convolutional layers and 1 pooling layer on the basis of VGG16

Images