隐式自校正广义预测控制器及其全局收敛性  

GLOBAL CONVERGENCE OF IMPLICIT SELF-TUNING GENERALIZED PREDICTIVE CONTROLLER

作　　者：舒迪前[1] 石中锁[1] 

机构地区：[1]北京科技大学自动化信息工程学院

出　　处：《自动化学报》1995年第5期545-554,共10页Acta Automatica Sinica

摘　　要：通过引入等价性能指标，给出了一种基于ＣＡＲＩＭＡ模型、考虑模型误差的隐式自校正广义预测控制器。并通过构造两个辨识器，使控制器的全部参数能递推估计。文末给出了当采用改进型最小二乘法和随机逼近法估计参数时的自校正广义预测控制隐式算法的全局收敛性证明。As to the convergency property of the implicit algorithm of the generalized predictive self-tuning controller, its satisfactory result has not yet been found. In this paper, by introducing an equivalent cost function and considering the parameter error of the model, an implicit self-tuning generalized predictive controller based on CARIMA model is presented. Parameters of the self-tuning controller can also be recursively estimated by introducing two identifiers. In the end, the convergent property of the proposed implicit algorithm is also given.

关 键 词：自校正控制 预测控制 全局收敛性 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置]