A new quantitative method for risk assessment of water inrush in karst tunnels based on variable weight function and improved cloud model

摘要

Water inrush disaster seriously affects the safety of karst tunnel construction. It's essential to assess the risk level of water inrush in karst tunnels accurately, and take some effective countermeasures to reduce the damage to the project. We integrates the variable weight theory and cloud model theory to construct the VW&ICM calculation model to evaluate the risk of construction of karst tunnels. First, we select the index factors that affect the risk of water inrush from karst tunnels. Secondly, according to the theory of variable weight, we construct the zoning variable weight model, and the normalization criterion of the data from index factors is established. Thirdly, according to the attribute mathematic theory, we select the numerical characteristics of the improved cloud model, and the normal cloud generator is used to establish the point membership function. Finally, the risk index data is brought into the variable weight vector function and the point membership function to obtain the variable weight vector W and the membership degree matrix R. The risk overall evaluation vector B is further calculated, and the membership degree fluctuation is comprehensively analyzed according to the principle of maximum membership degree. The scope determines the risk of water inrush from karst tunnels. Finally, calculate the variable weight vector W and the membership matrix R, and the overall risk evaluation vector B is further integrated by W and R. According to the max-subjection principle and the fluctuation rule of membership, we get the risk level of water inrush in karst tunnel. The VM&ICM not only can quantitatively evaluate risk by considering the uncertainty of risk assessment and the influence of indicator size on weight, but also can make an analysis of the reliability to make the assessment result more convincing. The model makes improvement in weakening the influence of subjective factors on assessment results and in allocating the weight of indicators. A simple and practical software package is developed, which greatly improves the computational efficiency of VW&ICM method. The VW&ICM calculation model is applied to the risk assessment of water inrush for karst tunnels, and the results are basically consistent with the on-site construction situation.