基于灰色预测与BP神经网络的极端降水事件预测

Prediction of Extreme Precipitation Events Based on Grey Prediction and BP Neural Network

摘要: 极端降水事件会对人们的生命、安全和财产构成严重的威胁，因此建立极端降水事件的预测模型是非常有必要的。本文采用灰色预测模型与BP神经网络两种方式对郑州、北京、太原的极端降水事件进行预测，绘制了相应图像，并且进行了误差分析。结果推断未来郑州降水量仍然偏高，北京未来可能会出现极端降水情况，太原发生极端降水的可能性很小。

Abstract: Extreme precipitation events pose serious threats to people’s lives, security and property. Therefore, it is necessary to establish a prediction model for extreme precipitation events. In this paper, the grey prediction model and BP neural network are used to predict the extreme precipitation events in Zhengzhou, Beijing and Taiyuan. The corresponding images are drawn and the statistical errors are analyzed. The results show that the precipitation in Zhengzhou is still high in the future, Beijing may have extreme precipitation in the future, and the possibility of extreme precipitation in Taiyuan is very small.