Intelligent control based on intelligent characteristic model and its application
Intelligent control based on intelligent characteristic model and its application
复制标题
DOI：
10.1360/03yf9019
复制
发表时间：
2003-06
期刊：
影响因子：
--
通讯作者：
Hongxin Wu;Yingchun Wang;Yan Xing
中科院分区：
文献类型：
--
作者：
Hongxin Wu;Yingchun Wang;Yan Xing
针对一类具有非线性和不确定性、被控输出变量不能连续在线测量的复杂对象，提出了一种基于智能特征模型的智能控制新方法。该方法的基本思想是根据被控对象的特性与控制要求相结合的原则，利用智能技术建立被控对象的特征模型，并在此基础上提出一种新的智能控制器设计方法。首先，给出了智能特征模型的建模原则和表达方法。然后在描述智能特性模型的基础上，给出了由多个具有定性和定量信息的开、闭环子控制器组成的智能控制器的设计原理和方法。最后介绍了该方法在真实的铝电解过程氧化铝浓度控制中的应用。实践证明，上述方法不仅易于工程实现，而且避免了一般智能控制器的反复试验。应用效果较好，实现了低铝浓度的长期稳定控制，阳极效应控制率由原来的60%大幅提高到80%。
This paper presents a new intelligent control method based on intelligent characteristic model for a kind of complicated plant with nonlinearities and uncertainties, whose controlled output variables cannot be measured on line continuously. The basic idea of this method is to utilize intelligent techniques to form the characteristic model of the controlled plant according to the principle of combining the characteristics of the plant with the control requirements, and then to present a new design method of intelligent controller based on this characteristic model. First, the modeling principles and expression of the intelligent characteristic model are presented. Then based on description of the intelligent characteristic model, the design principles and methods of the intelligent controller composed of several open-loops and closed-loops sub controllers with qualitative and quantitative information are given. Finally, the application of this method in alumina concentration control in the real aluminum electrolytic process is introduced. It is proved in practice that the above methods not only are easy to implement in engineering design but also avoid the trial-and-error of general intelligent controllers. It has taken better effect in the following application: achieving long-term stable control of low alumina concentration and increasing the controlled ratio of anode effect greatly from 60% to 80%.