采用 CARMH 模型的广义最小方差自校正控制器的全局收敛性

On the Global Convergence of the Generalized Minimum Variance SelfˉTuning Controller Based on a New Type Model

在一种新的预测模型的基础上，建立了一般的广义最小方差自校正控制器，并证明了当辨识算法采用随机逼近法时这种自适应算法的全局收敛性，从而为这种新型模型的进一步应用提供了一定的理论依据。

In this paper, based on a new calculation model, the generalized minimum variance self tuning controller is designed and the convergence and the stability of the self adaptive algorithm are proved when the cognitive algorithm takes the stochastic appoach, thus providing some theoretical basis for the further application of this new model \=