摘要
以绵阳市2014~2016年空气污染指数(API)以及SO 2、NO 2、PM 10等污染物为研究对象,探讨了绵阳市空气污染的变化规律,并分析它们与常规观测的地面气象资料之间的关系。尝试采用多元线性回归方法及BP神经网络方法建立污染预报模型,并检验分析两种模型的可行性。结果表明基于BP神经网络的预报模型在污染预报中可行,并建立基于BP神经网络进行空气质量预测的预测模型,利用历史资料进行验证。
In this paper,the air pollution index(API)and SO 2,NO 2,PM 10 pollutants in Mianyang City from 2014 to 2016 are taken as the research objects,and the variation law of air pollution in Mianyang City is discussed,and the relationship between them and ground meteorological data observed by conventional methods is analyzed.This paper attempts to use multiple linear regression method and BP neural network method to establish pollution prediction model,and test and analyze the feasibility of the two models.The results show that the prediction model based on BP neural network is feasible in pollution prediction,The prediction model of air quality based on BP neural network is established and verified by historical data.
作者
陶雪梅
李高
TAO Xue-mei;LI Gao(MianYang Meteorological Bureau,Mianyang,Sichuan 621000,China)
出处
《四川环境》
2020年第2期49-54,共6页
Sichuan Environment
关键词
气象因子
空气污染预报
多元回归
BP神经网络
Meteorological factors
air pollution prediction
multivariate regression
BP neural network unify with the Chinese abstract
