通感一体化架构下光纤振动系统的运行优化研究

Research on Operational Optimization of Fiber-Optic Vibration Systems under an Integrated Sensing and Communication Architecture

摘要: 长输管道沿线地质环境复杂，传统分布式光纤振动传感(DVS)系统在多源异构数据融合及环境噪声干扰方面面临显著工程瓶颈。对此，文章提出了一种基于通感一体化架构的系统优化方案：研究构建了“端–边–云”协同拓扑，设计了基于北斗授时的毫秒级时空对齐机制；针对光纤敷设余长导致的定位偏差及风沙温漂噪声，提出了基于GIS坐标的非线性余长修正方法，并设计了CNN-LSTM入侵识别模型。仿真实验表明，该方案有效抑制了复杂工况下的误报率波动，在包含农田灌溉、河流水位变化等干扰的实际环境下，系统平均误报率由初期的15.2%降至8.5%，验证了其在管道智能化防护中的有效性。

Abstract: The geological environment along long-distance pipelines is complex, and the engineering bottlenecks of traditional distributed fiber-optic vibration sensing (DVS) systems are in multi-source heterogeneous data fusion and environmental noise interference. In response, this paper proposes a system optimization scheme based on an integrated sensing and communication (ISAC) architecture. The study constructs a “terminal-edge-cloud” collaborative topology and designs a millisecond-level spatiotemporal alignment mechanism based on BeiDou time synchronization. To address the positioning deviation caused by fiber laying excess length and the noise from wind, sand, and temperature drift, a nonlinear excess length correction method based on GIS coordinates is proposed, along with a CNN-LSTM intrusion recognition model. Simulation experiments demonstrate that the proposed scheme effectively suppresses false alarm rate fluctuations under complex operating conditions. In real-world environments containing interferences such as farmland irrigation and river water level variations, the average false alarm rate of the system is reduced from an initial 15.2% to 8.5%, verifying its effectiveness in intelligent pipeline protection.