Remote fault detection and location of power fiber optic cable based on a logistic regression model

Abstract

Abstract The current fiber optic communication network presents the characteristics of network scale, complexity, and efficiency, and the use of traditional means for the maintenance of power fiber optic cable has been greatly limited. In this paper, based on the characteristics of simple structure, convenient construction, strong randomness, strong sequence autocorrelation, and good iterative performance of chaos theory in a logistic regression model, a logistic-based remote fault detection and location system for power fiber optic cables is proposed. The fault location test is carried out through with TMS200 series fiber optic cable automatic monitoring management system and GIS method. The fault location error of the logistic-based fault detection method is less than ±3m, that of the GIS method is less than ±10m, and that of the TMS200 series cable automatic monitoring and management system is less than ±15. The comparison shows that this paper’s fault detection and location system is more accurate and can help maintenance personnel have a faster and more comprehensive understanding of the line situation. Realize the rapid repair work of faults, which provides great help to maintain the fiber optic cable lines.