Self-Tuning Deconvolution Filter for Nonstationary Arma Signals
非平稳ARMA信号自校正去卷滤波器

Using the modern time series analysis,this paper presents a new self-tuning recursive deconvolution filter for unknown nonstationary ARMA input signal observed through a known linear system.It can be expressed by an ARMA innovation filter,and can be used to handle the nonminimum phase and unstable observation Systems.A Simulation example shows its effectiveness.