摘要
对于多变量系统,回路间的关联分析和变量配对是控制系统设计的第一步。文献针对稳态相对增益阵(relative gain array,RGA)只考虑了系统的稳态特性而没有考虑动态过程中各回路的影响的基础上提出了各种改进的动态相对增益阵。在多变量状态反馈预测控制(SFPC)的基础上提出了一种新的变量配对标准,能比较充分的反映控制过程的动态和稳态信息。通过对预测时域P的优化选择确定被控过程的相关性指数矩阵μ,并将μ与稳态信息阵K相结合得出最终的配对矩阵Λ。最后通过实例研究与其他配对方法比较,表明提出的方法能得出比较好的变量配对结果。
For the multivariable control system, interaction analysis and variable pairing are the first step for the control system design. In order to handle the problem that RGA do not reflect the dynamic characteristic among loops, all kinds of improved dynamic relative gain array are presented. The paper presents a new variable pairing method based on multivariable state feedback model predictive con trol, It can fully reflect the dynamic and stead state information about control process. Through the optimization of the prediction hori zon, the correlation index array can be identified. Combining the correlation index array and steady state array, the final pairing array is defined. Several cases and the comparisons with the existing methods indicate that the proposed method is a useful tool to give the best pairing scheme.
出处
《控制工程》
CSCD
北大核心
2013年第6期996-999,1004,共5页
Control Engineering of China
基金
国家自然科学基金资助项目(20976193)
关键词
多变量系统
变量配对
相对增益
状态反馈预测控制
muhivariable system
variable pairing
relative gain
state feedback predictive control
