Review of Slope Stability Analysis Methods: From Deterministic Analysis to Intelligent Prediction

Abstract

Slope stability analysis is a core topic in the field of geotechnical engineering, and its research methods are undergoing a profound transformation from deterministic analysis to intelligent prediction. This paper systematically reviews the latest research progress in slope stability analysis methods and clarifies the core contribution of this review in integrating traditional analytical frameworks with cutting-edge intelligent technologies for slope stability assessment. First, it elaborates on the deepened application of the limit equilibrium method in engineering practice and analyzes the advantages of numerical simulation techniques such as the finite element strength reduction method under complex geological conditions, as well as their complementarity with traditional methods. Second, focusing on special working conditions such as rainfall infiltration and seismic loads, it delves into the research progress on slope instability mechanisms under fluid-solid coupling, dynamic analysis, and the coupling of multiple factors. Third, it discusses reliability analysis methods that consider uncertainties in geotechnical parameters, including the application of random field theory and global sensitivity analysis in quantifying slope safety conditions. Finally, it highlights the cutting-edge applications of machine learning, deep learning, and intelligent optimization algorithms combined with new technologies such as UAV photogrammetry and 3D modeling in slope stability prediction, modeling of heterogeneous media, and data acquisition. The review points out that current research is trending toward uncertainty quantification, multi-method integration, and data-mechanics dual-driven development, and further identifies the key challenges in the field such as the physical interpretability of data-driven models and the scale-up from single-slope to regional analysis. In the future, physics-data fusion-driven methods, refined 3D heterogeneous modeling, and full life-cycle dynamic risk assessment are expected to become important development directions, which can provide direct technical guidance for engineering practice and disaster prevention in geotechnical engineering.