一类非线性系统盲辨识算法及仿真研究

Research of Algorithm and Simulation of a Class of Nonlinear Blind Identification

针对离散非线性系统盲辨识问题,提出了一种基于循环平稳输入的非线性系统盲辨识方法。利用循环平稳输入信号的一阶统计特性和Hammerstein-Wiener模型非线性部分的逆映射,将有输入信号的辨识过程转变为无输入信号的辨识过程。介绍了Hammerstein-Wiener模型的结构及循环平稳输入的统计特性,对盲辨识算法机理进行了阐述。仿真结果表明该方法在解决一类离散非线性系统盲辨识问题上的切实可行性。

The cyclostationary input with the Hammerstein-Wiener model was combined to realize the blind discrete nonlinear system identification. The first-order moment of the input and the inverse nonlinear mapping of the Hammerstein-Wiener model were introduced. The statistics characteristic of the input and the structure of the model were expatiated on, and then the mechanism of the algorithm was discussed. Simulation results demonstrate the effectiveness of this approach.