摘要
微细颗粒物危害人体健康,预测室内颗粒物浓度水平是分析室内颗粒物污染的特征规律、治理室内颗粒物污染的重要手段.作者在现有室内颗粒物机理模型的基础上,建立了多元回归模型和自回归滑动平均模型,并对室内PM2.5浓度水平进行了预测,通过比较分析两个模型的特征参数和统计结果,证明应用自回归滑动平均模型预测污染物浓度水平精确度更高.
Indoor fine particle pollution problem has become seriously recently because of the heavy traffic and factories discharge. To forecast indoor fine particle concentration is of im- portance and significance to analyze the pollution distribution and control the pollution emis- sion. In this paper, two different forecasting models are built up by multiple regression mod- el and ARIMA model, respectively. Both models＇ characteristic parameters and statistical re- sults were compared, and the predict results of the ARIMA model were verified to close the real value.
出处
《陕西科技大学学报（自然科学版）》
2008年第4期74-77,81,共5页
Journal of Shaanxi University of Science & Technology
基金
国家自然科学基金项目(编号:50408019)
霍英东教育基金会高等院校青年教师优选资助课题项目(编号:104006)
湖南省杰出青年科学基金项目(编号:06JJ1001)
