A novel multi-dimensional cloud model coupled with connection numbers theory for evaluation of slope stability

Abstract

Abstract Stability evaluation of a slope involves various fuzzy and correlation indicators randomly distributed in finite intervals. A novel multi-dimensional connection cloud model was presented here to address multiple uncertainties and distribution characteristics of indicators, and to depict the randomness and fuzziness of the measured index value belonging to the classification standard in the slope stability analysis. In the model, when simulating fuzzy and random characteristics of evaluation indicators in finite intervals, the numerical characteristics of connection cloud model were assigned on the basis of the analysis of identical-discrepancy-contrary (IDC) relationships between measured indicators and the classification standard to overcome the subjectivity. Considering the effect of indicator correlation in a unified way, the integrated connection degree of a grade was further specified for the evaluation sample. Moreover, case studies and comparisons of the proposed model with one-dimensional normal cloud model, extension model, and support vector machine (SVM) were performed to confirm the validity and reliability. The results indicate that this model employed to evaluate slope stability can clearly depict the random and fuzzy distribution features of measured data in finite intervals, and its calculation process is quicker and simpler than that of one-dimensional normal cloud model.