摘要
本文所述系统有两个特点:输入为不可量测的随机信号;系统可以具有不稳定的零点或者说系统可以是非最小相位的.文中研究了将一类输入信号与具有正确相位的脉冲响应从输出信号中分开的条件,并对以状态空间模型为基础的极大似然反褶积方法进行了改进.利用合成数据进行了仿真研究,仿真结果验证了非最小相位系统的可辨识性.
The system studied here is a non-minimum phase system whose input is an unmeasurable random signal. The condition is investigated, under which the input and the impulse response of a non-minimum phase system can be correctly extracted from the output. The maximumlikelihood deconvolution method based on state space model is improved in this paper. The identifiability of the non-minimum phase system is verified by the simulation results.
出处
《自动化学报》
EI
CSCD
北大核心
1989年第4期311-317,共7页
Acta Automatica Sinica
基金
国家教委科学基金资助的课题
关键词
系统辨识
参数估计
反褶积
相位可辨识性
信号处理
System identification
Parameter estimation
deconvolution
phase idetifiability
signal processing
