Design of embedded optical cable anti extemal damage early waming system based on lonely forest statistical algorithm

The buried cable laying method is gradually favored because it solves the environmental protection problems of urban construction. The new city basically chooses the buried type operation under the condition of pipeline permission, but it increases the maintenance difficulty for the distribution network operation and maintenance personnel. In order to meet the demand of this application, this paper designs a kind of external stress sensitive detection function of distributed sensing optical fiber, combined with the computing advantages of Fourier transform and lonely forest algorithm, transforms the time-domain information of optical fiber into frequency-domain information, and extracts the sensitive phase amplitude and position important parameters in the frequency domain, and compares them with the storage fault threshold value of power optical fiber external failure event In addition, the system can detect the failure event in time and send out warning information to avoid further cable damage. Compared with other data algorithms, the system detection and fault evaluation method designed in this paper has high accuracy, strong matching of application requirements and practicability to solve problems.