Abstract: Time series analysis model based on single monitoring point is independent which doesn’t fully consider the space correlation of similar monitoring points, and couldn’t reflect the overall trend and pattern of slope deformation. On the basis of single point grey forecasting model, we applied fuzzy clustering method to the time series relationship analysis of slope monitoring points and established a deformation prediction model in consideration of multipoint space relevance. The model is applied to slope engineering of Jinping hydropower station and the results show that the reliability and accuracy of this method are obviously higher than those of single point prediction model.

Key words: slope, time series, fuzzy clustering, multipoint monitoring, combinatorial prediction model