Research On Early Warning Methods For Faulty Of Military Direct Buried Optical Cable Lines

Military optical cable network is an important basic communication facility for country,but the faulty of optical cable line has always been the main cause of communication network failure.The direct buried optical cable line is the most popular line laying method for military optical cable lines,and its faulty-free operation has become the key to the safe operation of military optical cable network.As to the malfunction of military direct buried optical cable lines,the significance and importance of ex ante prevention is far greater than the after-the-fact repair.Therefore,this paper studies the early warning method for the faulty of military direct buried optical lines to effectively reduce the blocking rate of optical cable lines,improve the combat level and informatization degree of optical cable line maintenance units,and the combat communication support effectiveness of military optical cable network.The main research contents of the thesis are as follows:First,by analyzing the main causes of the faulty in the optical cable lines,from the perspective of the cable line itself and the external force damage,two kinds of early warning methods for faulty of military direct buried optical cable lines are studied by predicting the transmission loss and using the drone aerial photography.Secondly,the support vector regression model is applied to the prediction of transmission loss of optical cable lines.Then the variable weight combination prediction model is established.Combined with the analysis of the line transmission loss warning threshold,the faulty warning method of optical cable lines is studied from the perspective of transmission loss change.Thirdly,the necessity,work flow and key technologies of the UAV patrol method combined with the deep learning algorithm for the early warning of optical cable line faulty are studied.It is proposed to use the convolutional neural network to detect the the hidden danger of the external damage of the optical cable line in the aerial image of the drone.The simulation realizes the automatic identification and warning of the computer of engineering vehicles such as excavators,thus improving the working efficiency of the traditional optical cable patrol and inspection.Fourthly,based on MATLAB/GUI design,the line faulty early warning management system is realized,so that the early warning result can be presented visually,and combined with the relevant early warning report generated by the system,it provides a convenient tool for the army in the promotion and application of the early warning methods.