Research on Intelligent Disaster Prevention and Mitigation Method in High Flood Risk Area of River Basin Based on Artificial Neural Network

The continuous and rapid development of industrialization and urbanization in the past 40 years has caused the urban and rural landuse based on natural ecosystem in the western region of the Taihu Lake Basin to face the problems of landuse scale reduction and spatial structure fragmentation, which has led to a serious reduction in stormwater regulation and storage capacity. Starting from the effects of flood disaster mitigation in the river basin, this paper introduces the theory of complex adaptive systems, natural-based solutions, combined with the theoretical connotation of sponge city planning and construction, constructs a theoretical model of flood disaster complex system, and further builds a disaster mitigation effect evaluation model for sponge city planning and construction; the 3S method is used to evaluate and visualize the indicators, and the artificial intelligence BP neural network method is used to calculate the disaster reduction effect level of 357 plots in Yixing City. The results show that the disaster mitigation level in the mountain forest area and central urban area in the south of Yixing City is high, and the level in the north is low. This shows that the natural ecological large sponge has an obvious effect on urban flood disaster mitigation; appropriate LID measures can effectively compensate for the fewer defects of the large sponge; complementary, the through river network can reduce the flood disaster to a certain extent.