基于Markov状态切换水质时序自回归预测模型.pdfVIP

40 3 ( ) Vol. 40 No . 3 2010 5 Journal of Jilin University( Earth Science Edition) May 2010 基于M arkov 状态切换的水质时序自回归预测模型 牛军宜, 冯 平 , 300072 : 水质时序演变特征的研究及预测对制定合理可行的水污染防控措施有重要意 , 但水质时序的 结构复杂性和非平稳性是采用自回归模型进行预测的瓶颈针对上述问题, 作者将马尔可夫状态切换理 论( M arkov Sw itching) 应用于水质时序的自回归建模预测马尔可夫状态切换 自回归模型( MSAR) 是一 种研究具有变结构动力特征的时间序列分析方法, 对异方差时序有较强的适应性实例运用中, 首先对果 河桥断面的氨氮时序进行 BoxCox 变换, 然后运用M SAR 模型对其进行结构分析及预测结果表明: MSAR 模型能有效识别出该水质时间序列演变过程中的两种结构模式, 通过与经典自回归模型的预测精 度相比, 该方法的各项指标均优, 也说明该方法在水质时间序列动态结构分析和预测方面有良好的应用前 景 : 水质时序; 马尔可夫理论; 状态切换; BoxCox 变换; 自回归模型; 水污染 : X52 : A 文章编号: 1 715888(2010)030 5708 Water Quality Time Series Prediction Based on Markov Switching AutoRegression Model NIU Junyi, FEN G Ping S ch ool of Civ il Eng ineering , T ianj in Unive rsity , T ianj in 300072, China Abstract: Evolution features research and multistep prediction of water quality time series have important significance for making reasonable and applicable environmental protection measures. But the structural complexity and nonstationary is the bottleneck of autoregressive model application to w ater quality series prediction. Focusing on these problems, M arkov sw itching theory w as introduced to combined w ith autoregressive model for w ater quality series simulating and predicting. M arkov sw itching autoregressive model( M SAR) is suitable for the analysis of nonstationary time series w ith variable structure, and has a good adaptability to heteroscedastic time series. In the modeling the w ater quality time series of Guoheqiao section, the BoxCox transformation w as used to transform