IMPROVED CLOUD MODEL FOR STABILITY EVALUATION OF RESERVOIR SLOPES

The stability of reservoir slope is affected by various factors so that it is still difficult to accurately evaluate their stability. This paper overcomes the two existing problems:(1)the eigenvalues of cloud model are not unified and (2)the expected rather than actual membership degrees are used as the traditional normal cloud model method is actually applied. It proposes an improved normal cloud model. A program for model application is developed on MATLAB software platform. Twelve factors that mainly affect the reservoir slope stability are selected as the evaluation indexes. The comprehensive cloud model of each stability level in regard to every evaluation index is generated from the cloud generator. Meanwhile,the weight of every evaluation index are obtained by the analytic hierarchy process on the basis of improved entropy weight. The maximum average membership degree is determined as the stability grade of a slope. The proposed method is employed in the stability evaluation of 34 engineering examples of reservoir slopes. And the evaluation results are compared with the ones by the CSMR method recommended by the current national code as well as the actual state of these slopes. It shows that the accuracy of this method is higher than that of CSMR method,and it has good reliability and practicability. The method can also be referred when evaluating the stability of other rock slopes such as road,railway or building slopes.