一种非平稳随机信号模型的时变参数估计算法性能研究

An algorithm property analysis for time varying parameters estimation of nonstationary random signals model

采用递推最小二乘算法求解非平稳随机信号模型的时变参数．该方法的主要特点是计算量小，占用存贮空间少，没有矩阵求逆的问题．应用该方法对分段线性调频信号、多个线性调频信号及非线性调频信号进行分析。

The authors use recursive least squares algorithm to estimate time varying parameters of nonstationary random signals model. This method has the advantage of saving computation time and storage space, and does not require any matrix inversion. Finally, this paper analyze the effects of different basis time varying functions on parameters estimation by using this method for a piecewise signal with linearly time varing frequency, multi signal with linearly time varying frequencies, a single signal with non linearly varying frequency.