ABSTRACT
Due to the influence of many factors, there may exist long interrupt in the dam monitoring data, which is not suitable for the adaption of regression analysis. This paper introduced an improved gray model generalized from the GM(1, m) model to solve the problem of data interrupt in dam safety monitoring, and an improved gray model is applied to the fitting and prediction of the base displacement and stress data of a high arch dam. The results show that the residual modified non-equidistant MGM(1, m) model has high accuracy and can be applied to practical engineering.
