Abstract: Against the massive monitoring data of bridge-health monitoring system，bridge-health monitoring prediction model based on DBSCAN cluster analysis algorithm is put forward.According to segmenting primitive data files with cluster analysis for pretreatment，the algorithm effectively relieves the massive data storage problems.With regression algorithm，the data is further processed after pretreatment in order to complete the bridge-health prediction function.

Key words: bridge health monitoring, massive data processing, DBSCAN cluster algorithm, regression prediction

摘要： 针对桥梁健康监测系统的海量监测数据，提出了一种基于DBSCAN聚类分析算法的桥梁健康监测预测模型，该算法通过对原始数据文件分段进行聚类分析预处理，有效地缓解了海量数据的存储问题。进一步对预处理后的数据，采用回归拟合算法，完成对桥梁健康状况的预测功能。

关键词: 桥梁健康监测, 海量数据预处理, DBSCAN聚类算法, 回归预测