Distribution network insulator detection based on improved ant colony algorithm and deep learning for UAV.

Under the background of the accelerating speed of urban and rural construction, the geographical environment of overhead transmission lines has also changed greatly. Using unmanned aerial vehicle (UAV) to realize intelligent line inspection can significantly shorten inspection time and improve inspection efficiency. In this paper, the intelligent power inspection of UAVs is studied from two levels: path planning and UAV control, and the insulator is identified through actual image recognition. At the path planning level, the improved swarm intelligence algorithm is used to conduct simulation experiments on the UAV flight path to find a safe and effective route. Insulator identification and defect location of overhead transmission lines are trained on the insulator dataset collected by deep learning technology to achieve accurate insulator identification and improve the efficiency of UAV inspection, which has great application prospects in engineering.