为满足隧道信息化施工的需求，该文提出一种隧道拱顶沉降监测的多源数据融合方法。该方法结合多种隧道拱顶沉降监测技术的优势，采用加权最小二乘的融合算法构建隧道拱顶沉降参数模型，充分考虑不同沉降监测方法在各种干扰因素影响下的监测数据噪声，采用时间递推的方式获取监测数据的噪声方差，来确定权重分配比例。这一方法能够实时估计不同监测方法中的噪声方差及待测物理量的状态，并已成功应用于隧道拱顶沉降监测数据分析中。结果表明，该融合算法不仅模型构建简洁，而且在快速、准确估计待测物理量状态方面表现优异，数据融合的稳定性高，有效提升了融合数据的精度。

To meet the requirements of information-based construction in tunnels,a multi-source data fusion method for monitoring tunnel crown settlement has been proposed in this paper. This method integrates the advantages of various tunnel crown settlement monitoring technologies and employs a fusion algorithm of weighted least squares method to construct a tunnel crown settlement parameter model. By comprehensively considering the monitoring data noise under various interference factors from various settlement monitoring methods and the noise variance of the monitoring data is obtained through a time-recursive approach to determine the weight distribution proportion. The method enables real-time estimation the noise variance in different monitoring methods and the state of the target physical quantities and it has been successfully applied to tunnel crown settlement data monitoring. The results indicate that this fusion algorithm not only presents a simple model but also excels in quickly and accurately estimating the state of the target physical quantities,demonstrating high stability in data fusion and effectively improving the accuracy of the fused data.