小区域大气污染物预测研究

Prediction of Small Regional Atmospheric Pollutants

大气污染物预测对于预防地区污染具有重要的意义。小区域内的污染物快速预测,可以对区域内空气质量的微变化及时做出响应。采用神经网络建立了多种空气污染物的多输入的预测模型。分析表明,结合一种高关联度的污染物组成的双输入模型在预测大多数污染物时均达到较好的效果,是适合小区域内污染物预测的有效方法。

The prediction of atmospheric pollutants is of great significance to prevent regional pollution.The rapid prediction of pollutants in small areas may quick response to air quality micro-changes in the region.This paper presents a set of multiple pollutant modeling and prediction process established by neural network.Implementation shows that,the two-input models combining another highly correlated pollutant achieved good effect during most of the pollutants prediction,is effective method for pollutant prediction in small regions.