The deformation of landslides is a non-linear dynamic and complex process due to the impacts of both inherent and external factors. Understanding the basis of landslide deformation is essential to prevent damage to properties and losses of life. To forecast the landslides displacement, a hybrid machine learning model is proposed, in which the Variational Modal Decomposition (VMD) is implemented to decompose the measured total surface displacement into the trend and periodic components. The Double Exponential Smoothing algorithm (DES) and Extreme Learning Machine (ELM) were adopted to predict the trend and the periodic displacement, respectively. Particle Swarm Optimization (PSO) algorithm was selected to obtain the optimal ELM model. The proposed method and implementation procedures were illustrated by a step-like landslide in the Three Gorges Reservoir area. For comparison, Least Square Support Vector Machine (LSSVM) and Convolutional Neutral Network–Gated Recurrent Unit (CNN–GRU) were also conducted with the same dataset to forecast the periodic component. The application results show that DES-PSO-ELM outperformed the other two methods in landslide displacement prediction, with RMSE, MAE, MAPE, and R² values of 1.295mm, 0.998 mm, 0.008%, and 0.999, respectively.

This study investigated the deformation characteristics and mechanisms of the Baiyansizu landslide under the coupled effects of crack development, rainfall infiltration, and road loading. Numerical simulations were performed using GeoStudio software (Version 2018; Seequent, 2018) to analyze geological factors and external disturbances affecting landslide deformation and seepage dynamics. Four additional landslides (Tanjiawan, Bazimen, Tudiling, and Chengnan) were selected as comparative cases to investigate differences in deformation characteristics and mechanisms across these cases. The results demonstrate that rear-edge deformation of the Baiyansizu landslide was predominantly governed by rainfall patterns, with effective rainfall exhibiting a dual regulatory mechanism: long-term rainfall reduced shear strength through sustained infiltration-induced progressive creep, whereas short-term rainstorms generated step-like deformation via transient pore water pressure amplification. GeoStudio simulations further revealed multi-physics coupling mechanisms and nonlinear stability evolution controls. These findings highlight that rear-edge fissures substantially amplify rainfall infiltration efficiency, thereby establishing these features as the predominant deformation determinant. Road loading was observed to accelerate shallow landslide deformation, with stability coefficient threshold values triggering accelerated creep phases when thresholds were exceeded. Through comparative analysis of five typical landslide cases, it was demonstrated that interactions between geological factors and external disturbances resulted in distinct deformation characteristics and mechanisms. Variations in landslide thickness, crack evolution, road loading magnitudes, and rainfall infiltration characteristics were identified as critical factors influencing deformation patterns. This research provides significant empirical insights and theoretical frameworks for landslide monitoring and early warning system development.