Flood loss prediction of coastal city based on GM-ANN

Abstract

Flood loss prediction is very important in China. In this paper, flood factors of coastal city, such as geological sedimentation rate, rise in sea level height, precipitation, urban drainage pipe length, annual GDP and population throughout the year will be considered to predict flood loss. Firstly, AHP will be used to determine the weight of flood factors. Considering the characteristics of different factors, GM is applied to get predictive values of flood factors. Then predictive values and weights are applied to ANN method to obtain flood loss of coastal city. Finally, Shenzhen is regarded as an example to verify the feasibility of this methods GM, DGM and ANN methods compared.