本文考虑一类基于最小二乘估计和最小方差控制的自校正调节器的收敛性,这种调节器首先是由K.J.Astrom引进的.有关这种调节器的收敛性已有很多结果,但是从模拟中发现的发散性问题,还没有从理论上加以研究.本文着重研究自校正调节器的发散问题,同时也给出一个收敛性的结果. 考虑如下的ARMAX模型:y(t)=θ_0~Tφ(t-1)+b_1u(t-1)+e(t).这里θ_0和φ(t)分别是参数向量和数据向量,其定义如下 In this paper, we consider the convergence of a class of self-tuning regulators based on least square estimation and minimum variance control. This regulator was first introduced by KJ Astrom. There have been many results on the convergence of this regulator. However, The problem of divergence found in this paper has not been studied in theory.This paper focuses on the problem of divergence of the self-tuning regulator and gives a convergent result.We consider the following ARMAX model: y (t) = θ_0 ~ Tφ (t-1) + b_1u (t-1) + e (t) where θ_0 and φ (t) are the parameter vectors and the data vectors, respectively, and are defined as follows