BP神经网络在降水酸度预测中的应用

本文利用南昌市城市大气中SO2、NOX、TSP等浓度数据及降尘数据建立了BP神经网络的降雨酸度预测模型，结果表明：BP神经网络的预测模型不仅能较好地反映致酸因素与降水酸度的相互关系，而且预测精度也高于多元回归等模型。

Learning Process to Predict Acidity of Precipitation

Traditional acidyty of precipitation predicting techniques concentrate predominantly of multivariate regression models and uvivariable time-series models.These single mathematical function-based predicting techniques are unable to represent the relationship between dependent and independents as well as the neural networks by incorporating back-propagtion learning process.This research establishes a BP neural network to predict the acidity of precipitation ,which in volves four variables:the concentration of sulfur dioxide,nitrogen,total suspended particulate matter and cast.The results indicate that use a BP neural network outperforms regression models and time-series models in terms of predicting accuracy.