摘要
将多变量系统的Smith预估器参数的在线辨识方法的研究结果运用到多变量模糊RBF神经网络控制系统中 ,构造多变量模糊RBF神经网络Smith预估控制器 .详细地阐述了控制器的结构和设计方法 ,并对一个双输入双输出对象进行仿真研究 .仿真结果表明由此所构造的控制器具有良好的控制效果 ,且当对象的特性发生变化时 ,系统仍具有较好的辨识效果 ,控制器具有一定的鲁棒性和自适应能力 .
The online identification of the Smith estimator′s parameters is applied to the control system for the MIMO fuzzy RBF network and Smith controller for MIMO fuzzy RBF network is constructed. The structure of this controller and its design are described in detail. A simulation study for 2 input/2 outputs system is done. The result indicates that the controller remains high performances and has a satisfactory identification when characteristics of plant are changed. The result also shows that the controller possesses robust and adaptability.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2001年第9期31-33,共3页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家自然科学基金资助项目 (691 85 0 0 5 )
