一种含控制滞后的Wiener模型的辨识方法

IDENTIFICATION OF WIENER MODEL WITH TIME DELAY

研究了含控制滞后Ｗｉｅｎｅｒ模型的辨识．在一定条件下，这一模型满足一组回归函数．利用Ｆｏｕｒｉｅｒ级数部分和可以求得回归函数在若干点上的估计．基于这些估计值可以决定控制滞后步数，进而辨识线性子系统和非线性映射的参数．

In this paper, the identification of Wiener model with time delay is considered. Under some special conditions, Wiener model satisfies a group of regressive functions. With the help of Fourier series, values of regress functions in some points can be estimated. Based on these estimation values, the time delay constant is determined. Furthermore, it is easy to identify parameters of the linear and nonlinear system with a bi direction linear neural network. A simulation example is calculated to prove the effectiveness of the algorithm.