By using monitoring technology we can collect information of the dynamics of surrounding rock in the process of construction, and hence judging the stability of surrounding rock condition and determining the right time of secondary lining and verifying the rationality of supporting mode. Through the analysis of tunnel monitoring measurement data of Huaying mountain tunnel, we established several regression models for comparison, and obtained two regression models of high precision. Then we applied the optimum weighted combinatorial prediction model (OWCPM) to predict the arch crown settlement, and compared the result with those of single prediction models. The OWCPM is verified to be superior to single models. According to the results of the OWCPM, we analyzed the deformation rate, and hence determining the timing of secondary lining. The results show that the OWCPM in analyzing tunnel monitoring measurement data improves the prediction accuracy, and better reflects the development trend of crown settlement compared with single forecast models.