论文部分内容阅读
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 multivariable systems, inter-loop correlation analysis and variable pairing are the first steps in control system design. In the literature, various improved dynamic relative gain arrays are proposed based on the steady-state relative gain array (RGA) only considering the steady-state characteristics of the system without considering the influence of each loop in the dynamic process. Based on the multivariable state feedback predictive control (SFPC), a new variable matching standard is proposed, which can fully reflect the dynamic and steady state information of the control process. Determine the correlation exponential matrix μ of the controlled process by optimizing the selection of the predicted time-domain P, and combine μ with the stationary information matrix K to obtain the final matching matrix Λ. Finally, the case study is compared with other matching methods, which shows that the proposed method can obtain better matching result.