信号建模(2):双频率信号

Signal Modeling.Part B:Dual-Frequency Signals

针对标准倍频双频正弦信号和含有初始相位不同频率双频信号的建模问题,利用离散观测数据估计信号模型的特征参数,研究和提出了动态数据下梯度辨识方法、动态数据下随机梯度辨识方法、滑动数据窗多新息随机梯度辨识方法、递增数据长度下递推梯度辨识方法、递增数据长度下牛顿递推辨识方法等,给出了几个典型辨识算法的计算步骤。

For the dual-frequency sine signal modeling,based on the discrete observation data,this paper studies and presents the gradient identification methods,the stochastic gradient（SG）methods,the multi-innovation SG methods,the recursive gradient methods and the Newton recursive methods for estimating the characteristic parameters of signal models.Moreover,the computational steps of several typical identification methods are provided.