时变参数模型谱估计在导弹遥测信号特征提取中的应用

Time-dependent Model Parametric Spectrum Estimation for Characteristic Extraction of Missile Telemetry Signals

分别给出了TVAR和TVARMA模型参数估计及谱估计的推导过程,并探讨了TVAR和TVARMA谱估计方法在某导弹遥测速变信号处理中的应用.应用结果表明,时变参数模型谱估计能较好地克服交叉项和谱泄露问题,且谱的时频聚集特性更好,便于特征的提取.

The deduction procedures of parameter estimation and spectral estimation of the time-varied auto-regressive model and time-varied moving average auto-regressive model are given.Their application to telemetry signal characteristic extraction is studied.The above methods are used in experimental telemetry signals,and the results show that parametric model spectrum estimation can overcome the cross-term and spectrum disclosure and is more convenient for characteristic extraction.