Generalized Parameter Estimation Method for Model-Based Real‑Time Optimization

Zhang, Duo; Wang, Kexin; Zhang, Zhengjiang; Xu, Zuhua; Shao, Zili; Biegler, Lorenz T.; Tula, Anjan K.

doi:10.1016/j.ces.2022.117754

Cited by 6 publications

(8 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…cost fails to decrease. 4,10 Similar situations also occur in nonlinear optimization (see Sect. 2.2 in Reference 11).…”

Section: Introductionmentioning

confidence: 58%

“…The model is either directly convexified 7,8 or adapted properly in each iteration 9 to avoid oscillation. In the second case, the trial step calculated according to the mismatched model is too aggressive for the plant that the plant cost fails to decrease 4,10 . Similar situations also occur in nonlinear optimization (see Sect.…”

Section: Introductionmentioning

confidence: 77%

“…Algorithms that solve problem () repeatedly based on local models only guarantee upon‐convergence optimality. Inadequate models 6 and improper algorithm parameters 4 could lead to oscillation. To address this, References 23 and 24 proposed unconstrained RTO algorithms in the trust‐region framework.…”

Section: Preliminariesmentioning

confidence: 99%

“…Handling mismatches by reliable optimization algorithms has aroused great interest. In static model‐based optimization, structural plant‐model mismatch is successfully handled by algorithms based on the gradient matching idea, for example, integrated system optimization and parameter estimation, 2 modifier adaptation, 3 and generalized parameter estimation 4 . In each RTO iteration, the gradient of the plant is approximated, and the model is adapted so that its first‐order necessary condition of optimality matches that of the plant.…”

Section: Introductionmentioning

confidence: 99%

“…Therefore, the model needs to be adapted based on the local information around u k at each iteration. The model optimization problem(4) then uses this adapted model, where subscript k indicates the kth RTO iteration.…”

mentioning

confidence: 99%

See 4 more Smart Citations

A globally convergent composite‐step trust‐region framework for real‐time optimization

Zhang,

Li,

Wang

et al. 2023

AIChE Journal

Self Cite

View full text Add to dashboard Cite

Inaccurate models limit the performance of model‐based real‐time optimization (RTO) and even cause system instability. Therefore, a RTO framework that guarantees global convergence in the presence of plant‐model mismatch is desired. In this regard, the trust‐region framework is intuitive and simple to implement for unconstrained problems. Constrained RTO problems are converted to unconstrained ones by the penalty function, and global convergence is guaranteed if the penalty coefficient is large enough. However, a sufficiently large penalty coefficient is hard to determine and may lead to numerical difficulties. This paper addresses this issue and proposes a novel composite‐step trust‐region framework for constrained RTO problems that handles inequality constraints directly. The trial step is decomposed into a normal step that improves feasibility and a tangential step that reduces the cost function. A rigorous proof of its global convergence property is given.

show abstract

“…cost fails to decrease. 4,10 Similar situations also occur in nonlinear optimization (see Sect. 2.2 in Reference 11).…”

Section: Introductionmentioning

confidence: 58%