迭代模型Smith预估控制:算法和稳定性  

Iterative Model Smith Predictive Control: Algorithm and Stability

作　　者：齐朝阳[1] 郑恩让[1] 侯再恩[2] 

机构地区：[1]陕西科技大学电气与信息工程学院,陕西西安710021 [2]陕西科技大学理学院,陕西西安710021

出　　处：《控制工程》2015年第1期133-138,共6页Control Engineering of China

基　　金：陕西省教育厅项目(11JK0935);陕西科技大学博士科研启动基金(BJ-09)

摘　　要：为了解决Smith预估控制算法建立模型不精确的问题,并针对一类可重复运行的时滞过程,在模糊相轨迹模型Smith预估控制算法的基础上提出迭代模型Smith预估控制。该算法在不设计自适应的前提下,可以自适应得到精确的预估模型,同时辨识过程的时滞时间。证明指出只要原系统闭环稳定则迭代预估模型Smith预估控制系统稳定,且给出了算法的收敛性判据。仿真表明,所提方法不需要已知过程的模型,通过一定次数的"迭代"即可得到精确的预估模型,从而克服了常规算法控制品质依赖精确数学模型的缺陷。同时该算法具有较强的鲁棒性。In order to solve the Smith predictive control algorithm to model imprecise problems, and for a class of time delay process can be run repeatedly, this paper proposes a Smith iterative model control based on fuzzy phase trajectory model predictive control algorithm. Without designing adaptive law, an accurate predictive model algorithm can be obtained adaptively, while Identify the dead-time. The paper proved that the iterative prediction model Smith predictive control system would be stable as long as the original system is stable, and the convergence criteria is given. The simulation results showed that the proposed method doesn＇t require that the model is known. Through a certain number of iterations it can obtain a relatively precise predictive model, which can overcome dead-time and defects of the conventional algorithms control quality depending on accurate mathematical model. At the same time it is feasible with good robustness.

关 键 词：时滞过程 迭代模型 SMITH预估控制 稳定性和收敛性 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]