摘要

Landslide forecasting is one of the emerging frontier topics in geotechnical engineering because of its need to deal with both abrupt change and complexity. Based on previous investigation and our experience with Saito's method, an improved Saito curve was drawn with four stages corresponding to the displacement process of the position on the back of the landslide. The equations and curves of displacement, rate, and acceleration, along with the landslide time forecast model, were all based on the improved Saito method. The Verhulst-Grey model that is used to calculate the parameters of the landslide time forecast model was established on the basis of displacement monitoring data. Finally, the precision of the improved Saito method was verified using the landslide monitoring data of a mining slope and a natural slope. The results show that the improved Saito method is both reliable and economical to use for landslide forecasting.