鲁棒性多变量预测控制技术的研究和应用

Researches and Applications of Robust Multivariable Predictive Control

针对时变性强、强耦合和大时滞等过程控制问题,介绍多变量预测控制的一种应用-鲁棒性多变量预测控制技术(RMPCT)的特点及其模型算法,并且论述了预测控制在系统辨识的方面的应用;同时,对软测量仪表的计算误差进行了分析,并根据不同的原因建立了相应的校正模型。用先进控制可以极大地提高石油化工装置的技术水平和经济效益。而作为先进控制核心的系统辨识问题已成为控制系统研究的热点问题。

To the strong time-varying, strong coupling and time-delay, the application of a multi-variable predictive control, robust multivariable predictive control technology （RMPCT）, is introduced. The characteristics and model algorithms, and the predictive control system identification in the context of the application are discussed. At the same time, soft-sensing instrument calculation error is analyzed. The corresponding calibration model is set up according to different reasons. The advanced control can greatly improve technique and economic effectiveness of the petro-chemical units. As the core of advanced control system, the identification problem becomes the hot spot of control system research.