军用直埋光缆线路障碍预警系统的设计与实现

Design and implementation of obstacle early warning system for military directly-buried optial cable lines

军用直埋光缆线路的无障碍运行是军用光缆网安全运行的关键。通过研究分析导致军用直埋光缆线路障碍的两类主要原因,提出基于传输损耗预测和基于无人机航拍图像工程车辆目标检测的两种光缆线路障碍预警方法,并分别通过建立变权组合预测模型和采用深度学习Faster R-CNN目标检测算法得以实现。最后,基于Matlab GUI设计实现了军用直埋光缆线路障碍预警系统,并将其编译为可执行文件,扩展了使用范围,对我军光缆线路维护工作具有一定的参考意义。

The barrier⁃free operation of military directly⁃buried cable lines is the key to the safe operation of military optical cable networks.By the research and analysis of the two main reasons leading to obstacles in military directly⁃buried optical cable lines,two kinds of obstacle early warning methods of optical cable lines based on transmission loss prediction and object detection of engineering vehicle based on UAV aerial image are proposed,with which the Faster R⁃CNN(Region⁃CNN)object detection algorithm are realized by establishing variable weight combination prediction model and adopting deep learning respectively.Finally,on the basis of MatlabMatlab GUI(graphical user interface),the obstacle early warning system of military directly⁃buried optical cable lines is designed and realized,and compiled to be an executable file,expanding the scope of application.The research results have certain reference significance for the maintenance of military optical cable lines.