Recognition on the Relativity between Typhoon and Storm Surge Using Fuzzy Interval Information Matrix

摘要

This paper introduces the method of fuzzy interval information matrix to recognize the relation between typhoon and storm surge with monitor data in Chiwan and Beijiao in Guangdong, China. Comparison with the linear regression model, the new method has two advantages: (1) The root-mean-square error is less than that of the linear regress model; and (2) The new method can improve the recognizing accuracy of small samples effectively. The interval size of domain of the input and output, generally, influences the accuracy of the new method. An adaptive interval size may advance the recognizing accuracy between the wind speed of typhoon and storm surge. In this study, the recognizing relation between the wind speed and storm surge at stable monitoring position is discussed, which should be studied further in future research, such as introducing air pressure and radius of wind speed as parameter into information matrix.