Data-Driven Disturbance Rejection Predictive Control for SCR Denitrification System

详细信息    查看全文

• 作者：Xiao Wu ; Jiong Shen ; Shuanzhu Sun ; Yiguo Li ; Kwang Y. Lee
• 刊名：Industrial & Engineering Chemistry Research
• 出版年：2016
• 出版时间：May 25, 2016
• 年：2016
• 卷：55
• 期：20
• 页码：5923-5930
• 全文大小：456K
• 年卷期：0
• ISSN：1520-5045

文摘

This paper develops a data-driven disturbance rejection predictive controller (DRPC) for the selective catalytic reduction (SCR) denitrification system in a coal-fired power plant by using the technique of subspace identification (SID). First, to alleviate the modeling difficulties of the model predictive control (MPC), a subspace predictive controller is constructed from the input–output data of the system by directly adopting the subspace matrices as the predictor. Then, following the same subspace method, a disturbance observer (DOB) can also be designed, which estimates both the external disturbances and plant behavior variations. Unlike the conventional DOB-based control system, the disturbance estimation signal is fed back to the subspace predictive controller to improve the accuracy of the prediction, and the integrated DRPC is finally constructed. The resulting controller can remove the effect of unknown disturbances and modeling mismatches quickly while satisfying the input constraints in the optimization. Case studies demonstrate the advantage and effectiveness of the proposed approach.