人工神经网络在苏州空气污染预报中的应用  

Application of Artificial Neural Network to Air Pollution Prediction in Suzhou City

作　　者：白晓平[1] 张启明[1] 方栋[1] 洪维民[2] 刘峰磊[2] COSTABILE Francesca 王芬娟[1] 

机构地区：[1]清华大学核能与新能源研究院,北京100084 [2]苏州环境监测中心站,江苏苏州215004 [3]意大利国家研究委员会大气污染研究所

出　　处：《科技导报》2007年第3期45-49,共5页Science & Technology Review

摘　　要：人工神经网络在预测预报领域的应用越来越广泛。简单介绍了BP神经网络的基本原理,较详细地回顾了国内BP神经网络在空气污染预报领域的研究应用情况,并建立了苏州市区SO2浓度预报的BP神经网络,预报结果表明:该模型具有简便、快速、准确的优点,可推广用于其它空气污染物的预报。The artificial neural network has been widely used in the forecast and prediction due to the development of its theorectical basis and computer technology. The basic principle of BP neural network is briefly introduced, and the applications of BP neural network to air pollution prediction in China are reviewed in detail. It is applied to predict SO2 concentration in Suzhou, and it is shown that BP neural network is easy to implement and gives quite accurate results, and that it can also be used to predict other air pollutants such as PM10, NOxy and so on.

关 键 词：空气污染预报 工神经网络 BP神经网络 

分 类 号：X831[环境科学与工程—环境工程]