Grey Neural Network Method for Disaster Rank Prediction of Typhoon Storm Surge

Although the method of multi-index disaster rank prediction of typhoon storm surge can avoid missing too much disaster-causing information and is in favor of improving prediction precision, it faces the problem how to identify the key disaster-causing factors and develop a prediction model. To solve this problem, a multi-index grey neural networks prediction model is presented, which combines the identification method of the key disaster-causing factors by grey relational analysis with artificial neural networks (ANN) BP algorithm owning a better learning and training ability. In the end, case study based on the typhoon storm surge data in Qingdao area validates the feasibility and effectiveness of the proposed method.