基于CSA-RLS算法的Wiener模型辨识

Identification of Wiener Model with CSA-RLS Algorithm

Wiener模型由动态线性模块和静态非线性模块级联组成,广泛地应用于工业过程中。但对于带有中间噪声的Wiener模型的辨识研究少之又少,因此论文采用CSA-RLS算法对Wiener模型进行辨识。对于非线性模块用三次样条函数逼近,线性模块通过有限脉冲响应表示。最后再通过递推最小二乘算法进行参数辨识,而模型的定阶准则选用OVR和FOE方法。经数值仿真证明,利用CSA-RLS算法辨识参数的准确性相较于CSA-LS算法有所提高,且算法的收敛速度更快。

Wiener model consists of a dynamic linear block and a static nonlinear block.It is widely used in industrial process⁃es.However,there are few researches on the identification of Wiener models with internal noise.Therefore,CSA-RLS algorithm is used to realize the Wiener model identification.For its nonlinear block,the cubic spline function is applied to approximate the struc⁃ture.The linear block is represented by a finite impulse response.Finally,the parameter identification is carried out by recursive least square algorithm and the order of Wiener model is defined by OVR and FOE criteria.The numerical simulation shows that the accuracy of using CSA-RLS algorithm to identify parameters has improved greatly when compared with CSA-LS algorithm,and the convergence speed of CSA-RLS algorithm is faster.