Adaptive IMC using fuzzy neural networks for the control on non-linear systems

被引:0
|
作者
Sánchez, EG [1 ]
Izquierdo, JMC [1 ]
Bravo, MJA [1 ]
Dimitriadis, YA [1 ]
Coronado, JL [1 ]
机构
[1] Univ Valladolid, Sch Telecommun Engn, Valladolid, Spain
关键词
adaptive IMC; non linear MIMO plants; penicillin simulated plant; FasBack; fuzzy neural networks;
D O I
暂无
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
This paper introduces the use of FasBack neuro-fuzzy system for the identification and control of non linear MIMO plants within IMC scheme. FasBack presents fast stable learning guided by matching and error minimisation, and presents good MIMO identification performance. Emphasis is made on the on-line adaptive capability of FasBack that can be used to develop adaptive IMC strategies, which are of interest in the cases of badly learned dynamics or plant parameters varying with time. Results for the control of a theoretical non linear MIMO plant from the literature reveal satisfactory performance for static model and controller, while performance is improved if adaptation is enabled. In addition, the control of a simulated penicillin plant is studied under several realistic conditions, in which adaptation shows to improve performance.
引用
收藏
页码:792 / 801
页数:10
相关论文
共 50 条