针对多主体协作健康监测系统中光纤传感网络链路故障，提出了一种基于光开关和图论的光纤传感网络自修复方法。研究采用图论相关理论表述含光开关光纤传感器网络链路的连通情况，据此研究了光纤传感网络出现链路故障时光开关的切换策略，实现对失效光纤布拉格光栅（FBG）传感器信号的自修复。以航空铝板结构试验件为实验对象，针对光纤传感网络典型链路故障，对基于光开关和多主体协作的光纤传感网络自修复效果进行了对比实验。实验结果表明：在光开关和主体协作下，识别精度较没有修复时识别精度提高10.02 mm，仅比网络结构完好时识别精度降低3.61 mm，有效提高了载荷识别精度以及光纤传感网络的可靠性。

Regarding to the link failure of fiber optic sensor network in multi-agent collaboration health monitoring system, a self-healing method of fiber optic sensing network based on optical switch and graph theory is proposed. Using the graph theory, the connectivity of fiber optic sensor network link with optical switches is expressed, the switching strategy when link failures happen in fiber optic sensor network is studied, and the self-healing of failure fiber Bragg grating (FBG) sensor′ signals is realized. Using the aviation aluminum structure as the experimental object and aiming at the typical link failures of fiber optic sensor network, its self-healing effect based on optical switch and multi-agent collaboration through contrast experiment is studied. The experimental results show that, with optical switch and multi-agent collaboration, the recognition accuracy is improved by 10.02 mm compared with that of unrepaired model, which is only lower 3.61 mm compared with that of health model, showing that optical switch and multi-agent collaboration can effectively improve the loading recognition accuracy and the reliability of fiber optic sensor network.