多变量ARMA信号自校正最优去卷平滑估计方法及其渐近稳定性

Self-Tuning Optimal Deconvolution Smoothing Method for Multivariable ARMA Signal and Its Asymptotic Stability

本文讨论信号去卷估计问题．运用新息理论和射影方法，基于ＡＲＭＡ新息模型设计多变量ＡＲＭＡ信号最优会卷平滑器，讨论了平滑器的渐近稳定性．在信号模型及噪声统计未知时，通过在线辨识ＡＲＭＡ新息模型提出了ＡＲＭＡ信号自校正会卷平滑器．

This paper deals with the signal deconvolution estimation problem. Using the innovation theory, the optimal deconvolution smoother for mutivariable ARMA signal is presented. Also, the stability of the smoother is considered. By identifying ARMA innovation model, the self-tuning estindtoy is presented while the signal model and noise statistics are unknown.