Abstract:To explore the emergency traffic protection decision-making methods for the natural disasters on highways, a reasoning model based on case-based reasoning (CBR) and combined weighting method is proposed. By integrating the entropy weight method and expert scoring method, a comprehensive weight allocation mechanism for disaster characteristic indicators is constructed, and the weighted Euclidean distance algorithm is combined to optimize the accuracy of case similarity matching. Using the typical landslide cases as a historical database and comparing the landslide cases as the query objective, the effectiveness of the model is verified. The results show that the similarity between the landslide case and Case 2 in the case library is the highest (with a total similarity of 0.999 0). It is recommended to adopt the repair scheme of "anti-slide pile+drainage system+slope foot back pressure", which is highly consistent with the actual engineering logic. Compared with a single weighting method, the combined weighting method takes into account both data objectivity and expert experience. The study also points out that the performance of the model is limited by the completeness and parameter sensitivity of the case library. In the future, it is necessary to further strengthen the robustness through data standardization construction, dynamic optimization of disaster resistance coefficient, and cross validation of multiple models, which provides a propagable theoretical framework and technical path for the rapid response and scientific decision-making of landslide disasters.
