引用本文:王红军, 田铮, 党怀义.非线性时间序列建模的异方差混合双AR模型[J].控制理论与应用,2006,23(6):879~885.[点击复制]
WANG Hong-jun, TIAN Zheng , DANG Huai-yi.Heteroscedastic mixture double-autoregressive model for modeling nonlinear time series[J].Control Theory and Technology,2006,23(6):879~885.[点击复制]
非线性时间序列建模的异方差混合双AR模型
Heteroscedastic mixture double-autoregressive model for modeling nonlinear time series
摘要点击 1547  全文点击 1444  投稿时间：2005-07-28  修订日期：2005-12-08
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  
  2006,23(6):879-885
中文关键词  异方差混合双自回归模型  平稳性  BIC准则  ECM算法  非对称分布  多峰分布  条件异方差
英文关键词  heteroscedastic mixture double-autoregressive model  stationarity  Bayes information criterion  expectation conditional maximization (ECM)algorithm  asymmetric distribution  multimodal distribution  conditional heteroscedasticity
基金项目  国家自然科学基金资助项目(60375003); 国家航空基金资助项目(03153059).
作者单位
王红军, 田铮, 党怀义 西北工业大学理学院应用数学系, 陕西西安710072
中国科学院自动化研究所模式识别国家重点实验室, 北京100080
中国飞行试验研究院, 陕西西安710089 
中文摘要
      研究了可用于非线性时间序列建模的异方差混合双自回归模型(heteroscedastic mixture doubleautoregressive model, HMDAR), 给出了HMDAR模型的平稳性条件, 利用ECM (expectation conditional maximization) 算法来估计模型的参数, 运用BIC(Bayes information criterion) 准则来选择模型. HMDAR模型条件分布富于变化的特征使它能够对具有非对称或多峰分布的序列进行建模, 将HMDAR模型应用于几个模拟和实际数据集均得到了较为满意的结果, 特别是对波动较大的序列, HMDAR模型能比其他模型更好地捕捉到数据序列的特征.
英文摘要
      The heteroscedastic mixture double-autoregressive (HMDAR) model, which is designed to model nonlinear time series, is investigated in this paper. Firstly, the stationary conditions are derived. Secondly, the parameters are estimated via expectation conditional maximization (ECM) algorithm. Thirdly, the Bayes information criterion (BIC) is used to select the model. The varied feature of conditional distributions makes the HMDAR model capable of modeling time series with asymmetric or multimodal distribution. Finally, the model was applied to several simulated and real data sets with satisfactory results. Especially to a highly time-variant series, the HMDAR model shows better performance than other competing models in data capturing．