Affiliation:
1. Xi'an University of Technology
2. Sichuan University
Abstract
The controlled autoregressive model is widely used to describe a dynamic process. In this paper, a series of new algorithms are proposed to estimate the models coefficients and to predict future change of the process. On the one hand, it is can be proved that these new algorithms are outlier-tolerant in the case that there are outliers in sampling series. On the one hand, these new algorithms are near to the optimal estimators and predictors separately in normal case.
Publisher
Trans Tech Publications, Ltd.
Reference9 articles.
1. Box Q E P, Jenkins G M. Time Series Analysis: Forecasting andControl. San Francisco: Holden-day Press, (1970).
2. Hannan E J. The estimation of the order of an ARMA process. Annals of Statistics, 1980, 8(5): 1071-1081.
3. Boker J, Keviczky L. Structural properties and structure estimation of vector difference equations. International Journal of Control, 1982, 36(3): 461-476.
4. Deng Z L, Guo Y X. Dynamic System Analysis and Its Application. Shenyang: Liaoning Science & Technology Press, 1985: 31-130. (in Chinese).
5. Wei Wang, Junhong Li and Ruifeng Ding. Maximum likelihood parameter estimation algorithm for controlled autoregressive autoregressive models. International Journal of Computer Mathematics, 88(16): 3458-3467, (2011).