ABSTRACT
ABSTRACT: The effective way to forecast and evaluate the risk of urban water hazard has become a worldwide topic in the background of global warming and highly urbanized. A risk assessment method for urban water hazard is proposed according to the characteristics of coexistence of certainty with uncertainty in water disaster system. On the basis of the theory of neural networks and cloud model, four basic evaluation factors influencing the urban water disasters are selected, risk levels are determined according to the hydrological frequency curve and the comprehensive cloud model of all risk levels belonging to various evaluation factors are generated. Measured time series of evaluation factors are simulated and forecasted by RBF Artificial Neural Network (RBF-ANN) and distribution maps of degrees of certainty belonging to all risk levels are drawn according to predictions. Examples show that the method can improve the studying of the uncertainty among the ownership of the risk levels in the process of evaluation. Results can accurately reflect the risk levels of urban water disasters. Analyses show that the assessment method based on RBF-ANN-cloud model provides a new way of thinking for the prediction and evaluation of the urban water hazard.
