系统辨识 递推最小二乘辨识 matlab仿真实验

A. The selected model

The system model：

B. Identification algorithm used

1.Recursive least squares identification. 2.Recursive least square identification with forgetting factor： Take the forgetting factor

clear all
close all
clc
a=[1 -1.3 0.5]';b=[0.8 0.3]';d=3;  %实际系统参数与纯迟延
na=length(a)-1;nb=length(b)-1;  %%na,nb 为输出输入系数矩阵 A,B 的阶数
L=2000;%仿真长度
uk=zeros(d+nb,1);yk=zeros(na,1);  %输入输出初始化
yk_m=zeros(na,1); %模型输出
u=randn(L,1);  %输入信号，采用方差为1的白噪声序列
xi=sqrt(0.1)*randn(L,1);  %干扰信号，采用方差为0.1的白噪声序列

thetae_1=zeros(na+nb+1,1);  %参数初值
P=10^6*eye(na+nb+1