Risk evaluation of natural disasters based on connection function

There are many uncertain factors, such as stochastic, fuzzy and gray information in the risk analysis on natural hazard. The set pair analysis (SPA) deals effectively with the various uncertain factors contributing to evaluation of the risk level of natural disasters. The evaluation indicators and standards of natural disasters risk are analyzed by identity-discrepancy-contrary (IDC). The result, the connection numbers, still has uncertainty information. Thus, yielding the risk evaluation model of natural disasters based on connection function, which construct a set pair relation between the indicators of connection numbers and comprehensive evaluation standards, and describe uncertainty of the connection numbers by using connection function. The study showed that the proposed model takes into account only the uncertainty of risk evaluation indicator identification on natural disasters, also the uncertainty of result connection numbers. This approach gives full consideration to the uncertainty of systematic evaluation process along with the actual meaning of comprehensive evaluation functions. Therefore, this means it is able to reduce the uncertainty of final evaluation results and improve the accuracy and reasonability of evaluation results. This model is capable of reflecting actual situation of the risk evaluation on natural disaster affected by various uncertain factors and has a promotional value in the natural disaster risk assessment.