| 一类时变时间序列模型的参数估计研究 |
| |
| 引用本文： | 曾鹏,王绍棣. 一类时变时间序列模型的参数估计研究[J]. 南京邮电大学学报(自然科学版), 1998, 0(Z1) |
| |
| 作者姓名： | 曾鹏 王绍棣 |
| |
| 作者单位： | 南京邮电学院计算机科学与技术系 |
| |
| 摘 要： | 根据对一类时变时间序列模型结构特点的研究，提出了一种时变ＡＲ模型的递推参数估计算法。其原理是将时变参数的估计转化为对状态方程的状态估计，采用卡尔曼滤波推导出参数估计递推公式，并研究了其渐近稳定的充分条件。
|
| 关 键 词： | 时间序列模型，参数估计，状态方程 |
The New Parameter Estimation Algorithm of the Time Varying Time Series Model |
| |
| Zeng Peng Wang Shaodi. The New Parameter Estimation Algorithm of the Time Varying Time Series Model[J]. JJournal of Nanjing University of Posts and Telecommunications, 1998, 0(Z1) |
| |
| Authors: | Zeng Peng Wang Shaodi |
| |
| Affiliation: | Zeng Peng Wang Shaodi Department of Computer Science and Technology,Nanjing Institute of Posts and Telecommunications,210003,Nanjing,PRC |
| |
| Abstract: | According to the structure characteristics of a time varying time series model,a new recursive parameter estimation algorithm of the time varying AR model is proposed.Based on the transformation of the time varying parameter estimation to the state estimation for state equations,a recursive algorithm is derived according to the Kalman filter principle.The sufficient condition of asymptotical stability for the algorithm proposed is also given. |
| |
| Keywords: | Time series models Parameter estimation State equations |
| 本文献已被 CNKI 等数据库收录！ |
