Abstract: This paper aimed to enhance forecasting capabilities for geological hazards and pipeline safety. Prediction models for surface displacement of mountainous terrain and strain of buried pipelines were developed using genetic programming. Separate models,DX and DY,were created to predict horizontal and vertical surface displacement along the X and Y axes,respectively. Strain models for buried pipelines under tension and compression were categorized into L,U,and R directions. Real monitoring data were analyzed to determine model variables,parameters,and influencing factors. Genetic programming quantified the correlation between surface settlement and pipeline strain,fitting the optimal relationship through these models. The accuracy of the prediction models was assessed using the average root mean square error (RMSE) and the average absolute percentage error (MAPE). The results indicate that the two types of models established in this study have small errors in predicting mountain surface displacement and pipeline strain. The models demonstrate high accuracy and robustness,effectively capturing and addressing the nonlinear and variable challenges inherent in the deformation process. They perform well in extracting features and patterns,exhibit strong data adaptability,and contribute to improving the reliability and practicality of geological disaster warning.