Cloud model for stability evaluation of recently failed soil slopes based on weight inversion of influencing factors

The stability of newly failed soil slopes, particularly under the influence of heavy rainfall, presents a significant threat to the safety of on-site rescue personnel. It is urgent to find a quick and accurate method for evaluating the stability of such slopes. The evaluation methods used for reactivating ancient landslides cannot be directly applied because the shear strength of the soil at the sliding surface of ancient landslides improves over time. Common numerical analysis methods such as limit equilibrium slice method or finite element method require time-consuming on-site surveys, which may affect the progress of rescue operations. The cloud model evaluation method is suitable for evaluating the stability of these slopes as it has lower accuracy requirements for evaluation factors, thus compensating for the limitations of the aforementioned methods. However, the current research on the weight of evaluation factors in cloud model evaluation still has some deficiencies. Therefore, a weight inversion method is proposed to calculate the weights of each evaluation factor. Nine primary influencing factors, including slope height, which are easily obtainable, were selected as evaluation factors. By referencing the "Technical Requirements for Geological Hazard Investigation" and previous research findings, grading intervals for each evaluation factor were determined. The corresponding comprehensive cloud model was generated using the MATLAB programming platform. According to the value of each evaluation factor before sliding, the cloud model’s characteristic parameters for weight inversion were obtained. And the cloud model for evaluating the stability of the newly failed slope was established. The Python language and the Qt Designer tool were used to develop the application of stability evaluation. The stability of a newly failed slope in Lengshui Village, Yongchun County, Fujian Province, during November 3-8 2016, was assessed using this application. The result align closely with the actual on-site conditions, validating the accuracy of the proposed evaluation method. The efficient runtime of the application further demonstrates its speed.