基于BP模型的大气污染预报方法的研究

Study on Forecasting of Air Pollution Based on BP Model

BP模型是目前最为广泛应用的神经网络模型之一 ,它是一种简单而又非常有效的算法。笔者将BP网络模型引入到大气污染预报领域 ,并根据大气污染物含量与气象要素的关系建立了大气污染物含量的神经网络预报模型。计算结果表明 ,BP模型应用于大气污染预报具有较高的预测精度和良好的泛化能力 。

As one of the neural network models,BP model,which is a simple and effective algorithm,has been applied widely.BP model is applied by the authors in the field of air pollution forecasting,and a neural network forecasting model of air pollution is set up as well,based on the relationship between air pollutant contents and meteorological factors.The experimental computation results show that the BP model has good quality in terms of forecasting precision and generalization ability.Besides,it provides a new method of urban air pollution forecasting for the information society.