过程不确定性下丙烯精馏过程多变量预测控制技术应用

Application of Multivariable Predictive Control Technology in Propylene Distillation Process Considering Process Uncertainties

以丙烯精馏塔的多变量预测控制为研究对象,将卡尔曼(Kalman)滤波方法与动态反馈预测控制技术相结合,提出了一种带有积分输入补偿的Kalman滤波方法对系统不确定性干扰进行估计,然后将滤波后的输出、控制作用及状态信息动态反馈给多变量预测控制器以增强系统的抗干扰能力,提高控制系统的性能,并构建出丙烯精馏塔过程机理模型及仿真平台。仿真结果对比表明,采用本文改进的Kalman滤波方法使得多变量预测控制系统的控制性能和鲁棒性明显增强,生产过程更加平稳。

Taking multivariable predictive control of propylene distillation tower as research object,combining Kalman filtering method with dynamic feedback predictive control technology,a Kalman filtering algorithm with integral input compensation was proposed to estimate the uncertainties of the system. Then the filtered output,control action and state information were dynamically fed back to multivariable predictive controllers to enhance the resisting disturbance capacity of the system and to improve the performance of the system,and the process mechanism model and simulation platform of the propylene distillation tower were established. The simulation results show that the improved Kalman filtering method can improve the control performance and robustness of the multivariable predictive control system,and make the production process more stable.