宁夏大学学报（自然科学版）2012，Vol.33Issue(4)：416-419,4.
银川市空气污染指数的分析与预测
Analysis and Forecasting of Air Pollution Index in Yinchuan
摘要
Abstract
Air pollution index (API) is an effective way to assess air quality. On the basis of analysis change characteristics of API in Yinchuan, wavelet analysis and BP neural network are combined to set up "decomposition-forecast-reconstruction" model and "wavelet function replacement" model to forecast API of Yinchuan. Result shows the change characteristics of API in Yinchuan are that it decrease from 2001 to 2011 totally and fluctuate periodically from January to December; The outcome which is acquired through decomposing data using db10, forecasting and reconstructing are better than other wavelet; The forecast accuracy of "decompose-forecast-reconstruction" method is high and better than "wavelet function replace" model. So, it is an effective method to forecast API of Yinchuan.关键词
空气污染指数/小波分析/BP神经网络Key words
air pollution index/ wavelet analysis/ BP neural network分类
资源环境引用本文复制引用
李媛..银川市空气污染指数的分析与预测[J].宁夏大学学报（自然科学版）,2012,33(4):416-419,4.基金项目
国家自然科学基金资助项目(41261021) （41261021）
