水工建筑物安全监测中 ,异常值产生的主行为较复杂且不易判定。从灰关联理论出发 ,根据实际的观测资料 ,采用GM( 1,N)模型 ,结合较多相关信息 ,研究了对异常值进行有效判别和分析的处理方法。研究结果表明 ,灰关联模型GM( 1,N)可用于安全监测 ,而且可以有效发现变形 ,对分析变形的原因 ,具有一定的实际参考价值 . In the hydraulic construction safety monitoring, the main behavior caused by the abnormal value is complicated and difficult to judge. Based on the gray relational theory, the GM (1, N) model and the related information are used to deal with the effective discriminant and analysis of outliers based on the observed data. The results show that gray relational model GM (1, N) can be used for safety monitoring, and the deformation can be effectively detected. It has some practical reference value for analyzing the causes of deformation.