Risk Assessment for In‐Service Soil Excavated Slope Stability of Expressway in the Mountains

Stability risk assessment of in‐service slope is critical to the safety of mountain expressway transportation. Traditional stability risk assessment for in‐service slope often suffers from an incomplete assessment indicator system, weights that are significantly influenced by human subjectivity and remain static, and incompatibility issues when assigning indicator values to evaluation grade standards. To address these limitations, this study focuses on the maintenance and management of soil excavated slopes along mountainous expressways in Anhui Province. Based on the characteristics of in‐service slopes in such environments, existing research, and relevant normative documents, a hierarchical framework for stability risk assessment is established integrating risk induction, risk suppression, and hazard severity. Then, constant weights for the indicators are determined by integrating the Group Decision‐Making Analytic Hierarchy Process (Group‐AHP) and the Criteria Importance Through Intercriteria Correlation (CRITIC) method. Variable weight theory is then applied to synthesize dynamic weights for the evaluation indicators. The risk assessment is subsequently performed using the matter‐element extension method, and the applicability and rationality of the proposed model are validated through a case study. The results demonstrate that the variable weights determined by the improved AHP‐CRITIC‐variable weight theory not only mitigate the influence of a single decision‐maker inherent in subjective weighting methods and incorporate the role of objective data but also accentuate the importance of poorer‐performing indicators in the comprehensive evaluation. The accuracy rate in predicting the stability risk level of in‐service slopes reached 94.12%. Furthermore, the RMSE and Kendall's rank correlation coefficient of the proposed model were significantly superior to those of both the AHP‐variable weight model and the AHP‐CRITIC‐comprehensive weight model. Key sensitive indicators identified include the lowest monthly mean temperature, slope angle, vegetation coverage, and facility importance. This method demonstrates good practical value and offers a novel approach for assessing the stability risk of service soil‐excavated slopes along mountainous expressways.