TY - JOUR
T1 - PTFE 间歇聚合反应过程变周期动态经济优化控制
AU - Wang, Xuesong
AU - Zeng, Xiangyu
AU - Bo, Cuimei
AU - Tang, Shuqi
AU - Dong, Chao
AU - Li, Jun
AU - Zhang, Quanling
AU - Jin, Xiaoming
AU - Ye, Shengli
N1 - Publisher Copyright:
© 2022 Chemical Industry Press. All rights reserved.
PY - 2022/9/5
Y1 - 2022/9/5
N2 - Polytetrafluoroethylene (PTFE) batch polymerization production mode can meet the market demand of small batch, multi-purpose and high quality products. Aiming at the strong nonlinearity and large time delay in the PTFE polymerization process, a dynamic economic optimization control method based on free terminal is proposed in this paper. Firstly, the dynamic economic optimization problem is established by integrating the production cycle as a degree of freedom into the optimization variable. The control inputs are discretized into piece-wise sequences by using improved control variable parameterization method, and the dynamic economic optimization problem can be transformed into a nonlinear programming (NLP) problem. Then, the NLP problem is solved based on the penalty function method of gradient descent, and the rolling optimization control method in the free prediction time domain is used to optimize the control input and terminal time. Finally, compared with PI control and nonlinear model predictive control, the simulation results show that free terminal dynamic economic optimization control method has higher economic benefit and shorter production cycle, which highlights the flexibility of batch production.
AB - Polytetrafluoroethylene (PTFE) batch polymerization production mode can meet the market demand of small batch, multi-purpose and high quality products. Aiming at the strong nonlinearity and large time delay in the PTFE polymerization process, a dynamic economic optimization control method based on free terminal is proposed in this paper. Firstly, the dynamic economic optimization problem is established by integrating the production cycle as a degree of freedom into the optimization variable. The control inputs are discretized into piece-wise sequences by using improved control variable parameterization method, and the dynamic economic optimization problem can be transformed into a nonlinear programming (NLP) problem. Then, the NLP problem is solved based on the penalty function method of gradient descent, and the rolling optimization control method in the free prediction time domain is used to optimize the control input and terminal time. Finally, compared with PI control and nonlinear model predictive control, the simulation results show that free terminal dynamic economic optimization control method has higher economic benefit and shorter production cycle, which highlights the flexibility of batch production.
KW - PTFE
KW - batch polymerization
KW - dynamic real-time optimization
KW - economic optimization control
KW - free terminal
UR - http://www.scopus.com/inward/record.url?scp=85140144985&partnerID=8YFLogxK
U2 - 10.11949/0438-1157.20220454
DO - 10.11949/0438-1157.20220454
M3 - 文章
AN - SCOPUS:85140144985
SN - 0438-1157
VL - 73
SP - 3973
EP - 3982
JO - Huagong Xuebao/CIESC Journal
JF - Huagong Xuebao/CIESC Journal
IS - 9
ER -