空气重介流化床的基于逆系统方法的内模控制  

Fluidized bed internal model control based on inverse system method

作　　者：宋夫华[1] 李平[1] 

机构地区：[1]浙江大学工业控制国家重点实验室,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2006年第9期1540-1544,共5页Journal of Zhejiang University：Engineering Science

基　　金：国家"973"重点基础研究发展规划资助项目(2002C13312200)

摘　　要：为了提高大型空气重介流化床(ADMFB)的鲁棒性和抗干扰能力,提出了基于人工神经网络(ANN)逆系统方法的内模控制.在分析ADMFB系统可逆性的基础上,该方法利用ANN辨识得到的原系统逆模型与原系统相串连,运用逆系统方法的思想,将该多变量、非线性、强耦合的系统通过反馈线性化,解耦成多个相互独立的单输入单输出(SISO)的伪线性子系统,对求得的伪线性系统采用该方法进行控制.ADMFB的仿真结果表明,该方法不依赖于系统的精确数学模型,对恒值扰动具有良好的抑制能力,并当非线性参数发生变化或逆系统存在建模误差时,由该方法设计的闭环系统均具有良好的鲁棒稳定性.To improve the robustness and anti interference ability of large air dense medium fluidized bed （ADMFB）, an internal model control based on artificial neural network （ANN） inverse system method was proposed. On the basis of reversibility analysis of ADMFB, the inverse model approximated by ANN was cascaded with the original system in the method. Based on the idea of the inverse system method, the multivariable, nonlinear and strongly coupled system was decoupled into several independent single input and single output （SISO） pseudo-linear subsystems via feedback linearization. Then the obtained pseudolinear subsystem was controlled by the method. The simulation results of ADMFB show that the method does not depend on the accurate mathematical model of the system, has good restrain ability against constant disturbance. The closed system designed by this method can maintain well robustness stability when the system has variable nonlinear parameters and inverse modelling errors.

关 键 词：非线性内模控制 逆系统方法 神经网络 鲁棒稳定性 解耦 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]