论文部分内容阅读
本文提出了基于左右对角关联阵的多变量时变最小方差(combined left/right diagonal interactors multivariable time-variant minimum variance,CMTMV)多扰动系统性能评估方法.首先,将左右对角关联阵引入多变量系统,来表征输入输出时滞结构,根据不同扰动作用时间设计CMTMV控制器,结合多模型混合权重的思想设计多模型CMTMV控制器.然后,利用多模型CMTMV控制器得出各被控变量输出方差,将多模型CMTMV控制器下被控变量的平均方差作为性能评估的基准.最后,本文通过在数值仿真和重油分馏过程中的控制系统性能评估的应用,验证了CMTMV评估方法的有效性.
In this paper, we propose a method to evaluate the performance of the multi-perturbation time-variant minimum variance (CMTMV) multi-perturbation system based on the left and right diagonal associated matrices.Firstly, the left and right diagonal correlation matrices are introduced into multivariable System to characterize the input-output time-delay structure and design the CMTMV controller according to the different turbulent action time, and design a multi-model CMTMV controller based on the idea of multi-model mixing weight.Then, using the multi-model CMTMV controller, the output variance of each controlled variable , The average variance of the controlled variables under the multi-model CMTMV controller is taken as the benchmark for performance evaluation.Finally, the effectiveness of the CMTMV assessment method is verified through the application of performance evaluation of control system in numerical simulation and heavy oil fractionation.