Rainfall station position optimization method based on rainfall analysis

The invention relates to the technical field of rainfall measurement, in particular to a rainfall station position optimization method based on rainfall analysis. According to the method, all the rainfall stations are divided into a plurality of rainfall sub-regions according to whether all the rainfall stations have rainfall, light rain and heavy rain in the same time period, and the rainfall stations in each rainfall sub-region have high rainfall generality; for rainfall stations with the same rainfall generality, a neural network model is adopted to fit a coordinate change rule and a rainfall change rule, the rainfall of a region without the rainfall stations can be calculated through the neural network model, and then the equal difference change of the rainfall of the adjacent rainfall stations is taken as an optimization target. Adjusting the position of a rainfall station in each area by adopting a particle swarm algorithm; due to the fact that adjustment is carried out on the original basis, it can be