A Study on Evaluation Method of Tropical Cyclone Disaster Assessment Based on Artificial Intelligence Technology:Taking Guangdong Province as an Example

The spatial distribution characteristics of tropical cyclone (TC) rainfall and direct economic losses incurred by TCs in Guangdong Province are systematically studied based on historical observation data. The results show that the coastal areas of western Guangdong are most frequently affected by TC winds, whereas the economic loss rates and disaster frequencies are much larger in the southwestern and eastern seashore areas of Guangdong. According to the spatial distribution characteristics of TC precipitation, high winds and rate of economic loss, Guangdong Province is divided into four regions. The evolutionary modeling method is then used to build the assessment model of TC economic loss rate in each region. The correlation coefficient between the simulated TC economic loss rate and the actual value is larger than 0.66, and the correlation coefficient between the pre-evaluation results and the actual values of independent samples reaches above 0.61 (significant at the level of a=0.05). This result shows that the regional assessment model based on the evolutionary modeling method has potential application value in the assessment of TC disaster in Guangdong Province.