基于BP神经网络的降雨量预测研究

Study on Rainfall Forecast Based on BP Neural Network

使用2种方法,即用前5年降雨量来预测后1年的降雨量和用气象资料预测降雨量,应用BP神经网络对降雨量进行预测,试图找出应用BP神经网络预测降雨的最佳方法。通过对青海省10个气象站的降雨进行预测分析,得到结论如下:两种方法均适合于降雨预测,其被选择概率均等。

To find the best method applying BP neural network to forecast rainfall,two methods with BP neural network are used to forecast rainfall in this paper.One is using the rainfall data of former 5 years to predict the rainfall of following 1 year.The other is using meteorological data to predict rainfall.Through forecasting and analyzing the rainfall in 10 meteorological stations of Qinghai Province,it is concluded that both of the two methods are suitable for rainfall forecast,both of them are chosen equiprobably.