基于BP网络的城市空气质量中长期预测研究  

Study on the Middle-long Term Prediction of City Air Quality by BP Network

作　　者：王艳平[1,2] 周广柱[3] 谢正苗[2] 

机构地区：[1]杭州电子科技大学管理学院,杭州310018 [2]杭州电子科技大学环境科学与工程研究所,杭州310018 [3]山东科技大学化学与环境工程学院,青岛266510

出　　处：《科技通报》2010年第2期245-248,共4页Bulletin of Science and Technology

基　　金：山东省软科学项目(2007RK071)资助

摘　　要：采用Matlab中BP神经网络的工具箱函数,对济南、青岛两市空气质量中长期变化趋势进行的预测显示,网络的预测精度高,自适应性强,训练速度快,避免复杂运算,节省大量时间,预测效果理想。因此,利用BP网络对空气质量的预测结果可以为环境管理部门的规划、管理工作提供依据,为保障人民的健康和社会生活服务。To predict the middle-long term trend of air quality in Ji ＇nan and Qingdao,the BP neural＇s toolox functions in Matlab had been introduced. The prediction result showed that the precision of the prediction is quite higher than other prediction models. BP neural network adapted to the city prediction easily,and the training speed was very fast,which will save a lot of time and avoid complex caculations. Therefore, BP neural network is a nice choice to predict the air quality of a city. Aeeoding to the prediction outcomes of air quality ,the environmental management departments will work conveniently and accurately ,which will guarantee a healthy and comfortable life for the citizens.

关 键 词：BP神经网络 空气质量 预测 中长期 

分 类 号：TP23[自动化与计算机技术—检测技术与自动化装置] TP29[自动化与计算机技术—控制科学与工程]