非线性时间序列门限自回归模型在环境空气质量预报中的应用  

An Application of Threshold Auto-Regressive Model to Predicting Ambient Air Quality

作　　者：潘磊 沙斐 

机构地区：[1]上海市浦东新区环境监测站

出　　处：《上海环境科学》2007年第5期212-214,218,共4页Shanghai Environmental Sciences

摘　　要：门限自回归模型作为1种利用历史资料对非线性时间序列进行描述从而进行预测的数学模型，多年来在各种统计领域中均取得良好的效果。根据非线性时间序列的门限自回归模型（简称TAR）的基本思路，利用环境空气自动监测系统历史监测数据资料，建立了浦东新区环境空气质量的预报计算模型。通过统计检验，检查使用TAR模型预测浦东新区环境空气质量与实际监测情况的符合程度。讨论了该模型在监测工作中的应用可行性。The threshold auto-regressive （TAR） model is an mathematical one that uses historical records to describe the non-linear time series for prediction, and has revealed very effective in various fields of statistics. Such a model to predict the ambient air quality in Pudong New Area of Shanghai was set up. Through statistical tests, the air quality in Pudong New Area predicted by using the TAR model was checked with the actual observations. The feasibility of its application in the environmental monitoring was then discussed.

关 键 词：环境空气质量 门限自回归模型 统计检验 预报 

分 类 号：X823[环境科学与工程—环境工程] O212.1[理学—概率论与数理统计]