arxiv: 2605.06469 · v1 · submitted 2026-05-07 · 🧮 math.OC · cs.LG· cs.SY· eess.SY

Dynamic Controlled Variables Based Dynamic Self-Optimizing Control

Chenchen Zhou , Shaoqi Wang , Hongxin Su , Xinhui Tang , Yi Cao , Shuang-hua Yang This is my paper

Pith reviewed 2026-05-08 08:07 UTC · model grok-4.3

classification 🧮 math.OC cs.LGcs.SYeess.SY

keywords dynamic self-optimizing controldynamic controlled variablesimplicit control policydata-driven controldeep neural networksbatch process optimizationvariable-horizon control

0 comments

The pith

Dynamic controlled variables held constant achieve near-optimal performance in variable-horizon processes where explicit policies fail.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper extends self-optimizing control to dynamic problems such as batch operations and grade transitions. It defines dynamic controlled variables as quantities that, when held constant, translate dynamic optimization into a simpler control task. A data-driven method uses deep neural networks to identify these variables as an implicit policy. Theoretical analysis shows this policy is more general than explicit control strategies and works for non-fixed horizons. Case studies confirm it handles multi-valued and discontinuous mappings that defeat traditional approaches.

Core claim

The central claim is that dynamic controlled variables (DCVs) form an implicit control policy obtained by mapping process states to quantities whose constant values approximate the solution of dynamic optimization problems. This mapping, parameterized by deep neural networks, is shown to be more general than explicit control laws and directly applicable to processes whose time horizons are not fixed in advance.

What carries the argument

Dynamic controlled variables (DCVs), a state-to-variable mapping learned by deep neural networks that is held constant to realize dynamic self-optimization.

If this is right

DCVs can represent multi-valued and discontinuous optimal mappings that explicit controllers cannot capture.
Self-optimizing control becomes feasible for batch and semi-batch processes whose horizons are not predetermined.
The implicit policy relates to conventional feedback controllers yet removes the need to re-solve the dynamic optimization at each horizon change.
A single neural-network training step replaces repeated explicit controller redesign for different operating conditions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same DCV construction could be inserted inside model-predictive control loops to reduce the frequency of full re-optimization.
Because the mapping is learned from data, the method may transfer to domains outside chemical engineering such as variable-length robotic task scheduling.
If the neural network is trained online, the approach could adapt to slow drifts in process parameters without full re-identification.

Load-bearing premise

Keeping the learned dynamic controlled variables exactly constant produces near-optimal trajectories even when the process horizon varies.

What would settle it

A batch reactor or grade-transition simulation in which holding the DCVs constant produces economic loss more than a few percent above the loss obtained by repeated dynamic optimization on the same instances would falsify the claim.

Figures

Figures reproduced from arXiv: 2605.06469 by Chenchen Zhou, Hongxin Su, Shaoqi Wang, Shuang-hua Yang, Xinhui Tang, Yi Cao.

**Figure 1.** Figure 1: Relationship between these three forms of DCV view at source ↗

**Figure 2.** Figure 2: Online implementation structure of DCV Theorem 1. Let F : R m ⇒ Y ⊆ R n be a set-valued mapping with closed graph, defined as gph F = {(x, y) ∈ R m × Y | y ∈ F(x)} and F(x) ̸= ∅ for all x ∈ R m. Then there exists a continuous function g : R m × R n → R n such that the solution mapping S : R m ⇒ Y, S(x) = {y ∈ R n | g(x, y) = 0} satisfies S(x) ̸= ∅ for all x ∈ R m and F(x) = S(x), ∀x ∈ R m. 14 view at source ↗

**Figure 3.** Figure 3: Composition of a sample In the face of uncertainties, the performance of DCVs necessitates a more nuanced evaluation. One approach leverages the worst-case loss, while another harnesses the statistical properties of the loss distribution, such as the expected loss over all uncertainties. In this paper, without loss of generality, further assuming the controlled variable is consistently maintained at 0 in… view at source ↗

**Figure 4.** Figure 4: Comparison of two loss function Bellman’s optimality principle, then solving a fixed finite-horizon problem. Model predictive control and approximate dynamic programming exemplify such methods. For problems with free terminal time, the switching or final time is usually introduced as a decision variable by input parametrization, transforming the problem into a nonlinear program. Current research on dynamic… view at source ↗

**Figure 5.** Figure 5: Comparison between implicit and explicit form view at source ↗

**Figure 6.** Figure 6: Closed-loop simulation under −20% disturbance constraint violation relative to the constraint upper bound across all scenarios, computed as: 1 N X N i=1 max(max t (gi(t)) − gmax, 0) gmax ,where gi(t) denotes the constraint value at the t-th instant in the i-th scenario, gmax stands for the upper limit of the constraint. The maximum violation value signifies the maximum constraint violation across all distu… view at source ↗

**Figure 7.** Figure 7: Multimodal solution case 3 modal solutions are obtained by solving the optimal control problem with multiple initial values. A total of 1600 globally optimal solutions and 1115 locally optimal solutions were obtained in 1600 disturbance scenarios. For training the control policy, 500 scenarios with local optimal solutions were randomly selected, and the data for those disturbance scenarios were removed fro… view at source ↗

**Figure 8.** Figure 8: Simulation results in Case 4 multimodals view at source ↗

read the original abstract

Self-optimizing control is a strategy for selecting controlled variables, where the economic objective guides the selection and design of controlled variables, with the expectation that maintaining the controlled variables at constant values can achieve optimization effects, translating the process optimization problem into a process control problem. Currently, self-optimizing control is widely applied to steady-state optimization problems. However, the development of process systems exhibits a trend towards refinement, highlighting the importance of optimizing dynamic processes such as batch processes and grade transitions. This paper formally introduces the self-optimizing control problem for dynamic optimization, termed the dynamic self-optimizing control problem, extending the original definition of self-optimizing control. A novel concept, "dynamic controlled variables" (DCVs), is proposed, and an implicit control policy is presented based on this concept. The paper theoretically analyzes the advantages and generality of DCVs compared to explicit control strategies and elucidates the relationship between DCVs and traditional controllers. Moreover, this paper puts forth a data-driven approach to designing self-optimizing DCVs, which considers DCV design as a mapping identification problem and employs deep neural networks to parameterize the variables. Three case studies validate the efficacy and superiority of DCVs in approximating multi-valued and discontinuous functions, as well as their application to dynamic optimization problems with non-fixed horizons, which traditional self-optimizing control methods are unable to address.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper extends self-optimizing control to dynamic regimes by defining dynamic controlled variables that are kept constant via an implicit policy, with a DNN used to learn the mapping for non-fixed-horizon problems.

read the letter

The core contribution is the shift from steady-state self-optimizing control to a dynamic version. They introduce dynamic controlled variables (DCVs) that are not fixed setpoints but state-dependent quantities held constant to achieve near-optimal dynamic performance. An implicit policy replaces explicit control laws, and they frame DCV design as a mapping identification task solved with deep neural networks. The theoretical section compares this to explicit strategies and links it back to conventional controllers, while the case studies target multi-valued functions, discontinuities, and variable-horizon problems that standard SOC cannot handle directly. That last point is the practical hook for batch and transition processes. The work is coherent on its own terms and fills a clear gap in the process-control literature. The data-driven route is presented cleanly as an independent step rather than circular. Soft spots are mainly in the execution details that the abstract leaves out. The claim that constant DCVs deliver near-optimality rests on the DNN mapping being accurate and generalizable; without reported error metrics, training-set coverage, or out-of-sample tests it is hard to judge how robust the method is in practice. The theoretical advantages are asserted but would need the actual derivations to confirm they do not rest on overly restrictive assumptions about the underlying model. For readers already working in self-optimizing control or dynamic real-time optimization this is worth a look, especially if they need to handle non-fixed horizons. It is not a broad methodological advance but a targeted, usable extension. A serious editor should send it to referees because the idea is well-motivated and the case studies address the stated limitations, even though revisions will be needed to supply quantitative validation and fuller proofs.

Referee Report

1 major / 4 minor

Summary. The paper extends self-optimizing control from steady-state to dynamic optimization problems by introducing the concept of dynamic controlled variables (DCVs) and an associated implicit control policy. It provides a theoretical analysis of the advantages and generality of DCVs relative to explicit control strategies, clarifies their relationship to traditional controllers, and proposes a data-driven design method that frames DCV selection as a mapping identification task solved via deep neural networks. Three case studies are presented to demonstrate the approach on multi-valued and discontinuous functions as well as dynamic problems with non-fixed horizons.

Significance. If the theoretical analysis is rigorous and the case studies provide clear quantitative evidence of superiority, the work could meaningfully advance self-optimizing control for dynamic processes such as batch operations and grade transitions. The data-driven DNN parameterization is a notable strength, as it directly addresses limitations of fixed-horizon explicit methods and enables handling of complex mappings without requiring explicit process models.

major comments (1)

[Theoretical Analysis] In the theoretical analysis section, the claim that maintaining DCVs at constant values achieves near-optimal dynamic performance rests on the accuracy of the DNN mapping, yet no error bounds or sensitivity analysis with respect to approximation error is provided; this is load-bearing for the generality argument.

minor comments (4)

[Data-Driven Design] The loss function and training procedure for the DNN in the data-driven design method are not specified, hindering reproducibility of the mapping identification step.
[Case Studies] Case study 3 on non-fixed horizons reports qualitative superiority but omits quantitative metrics such as economic loss or convergence rates compared to baselines.
[Problem Formulation] Notation for DCVs and the implicit policy should be introduced with explicit mathematical definitions in the problem formulation section rather than later.
[Figures] Figure captions in the case studies lack sufficient detail on axes, legends, and parameter values used in the simulations.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive review and the recommendation for minor revision. We address the single major comment below.

read point-by-point responses

Referee: [Theoretical Analysis] In the theoretical analysis section, the claim that maintaining DCVs at constant values achieves near-optimal dynamic performance rests on the accuracy of the DNN mapping, yet no error bounds or sensitivity analysis with respect to approximation error is provided; this is load-bearing for the generality argument.

Authors: We agree that the near-optimality claim in the theoretical analysis depends on the quality of the learned DCV mapping and that the absence of explicit error bounds or sensitivity analysis constitutes a limitation for rigorously supporting the generality argument. The manuscript establishes that an exact implicit mapping yields optimality by construction and invokes the universal approximation property of DNNs to justify the parameterization, but it does not quantify the effect of residual approximation error on closed-loop dynamic performance. In the revised version we will add a dedicated paragraph in the theoretical analysis section that (i) relates the training loss directly to the sub-optimality gap and (ii) provides a qualitative sensitivity discussion based on the uniform continuity of the optimal value function with respect to the controlled-variable deviation. This addition will be limited to the existing theoretical framework and will not require new theorems. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained

full rationale

The paper extends the definition of self-optimizing control to dynamic problems, introduces DCVs as a new concept, presents an implicit policy, and treats DCV design as an independent mapping-identification task solved via DNN parameterization. Theoretical comparisons to explicit strategies and case studies on multi-valued functions, discontinuities, and non-fixed horizons are presented as separate validation steps. No load-bearing step reduces by construction to fitted inputs, self-citations, or renamed prior results; the constant-value property is an assumption whose validity is checked externally via the DNN accuracy and case studies.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claim rests on the unproven assumption that constant DCV setpoints yield near-optimal dynamic trajectories and on the empirical success of DNNs to learn the required implicit mappings. No explicit free parameters or invented physical entities are named in the abstract.

axioms (1)

domain assumption Maintaining dynamic controlled variables at constant values achieves near-optimal performance for dynamic processes
Stated as the core expectation of the dynamic self-optimizing control definition

invented entities (1)

Dynamic controlled variables (DCVs) no independent evidence
purpose: Implicit policy variables whose constant setpoints approximate dynamic optima
Newly introduced concept whose independent evidence is the claimed theoretical analysis and case studies

pith-pipeline@v0.9.0 · 5570 in / 1117 out tokens · 17085 ms · 2026-05-08T08:07:20.155137+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

72 extracted references

[1]

Annaswamy, Karl H

Anuradha M. Annaswamy, Karl H. Johansson, and George J. Pappas.Control for Societal-Scale Challenges: Road Map 2030. IEEE Control Systems Society Publication, May 2023. 44

2030
[2]

A. Foss. Critique of chemical process control theory.IEEE Transactions on Automatic Control, 18(6):642–652, December 1973

1973
[3]

Control structure design for complete chemical plants.Computers & Chemical Engineering, 28(1):219–234, January 2004

Sigurd Skogestad. Control structure design for complete chemical plants.Computers & Chemical Engineering, 28(1):219–234, January 2004

2004
[4]

john Wiley & sons, 2005

Sigurd Skogestad and Ian Postlethwaite.Multivariable Feedback Control: Analysis and Design. john Wiley & sons, 2005

2005
[5]

Perspectives on the synthesis of plant-wide control structures.Journal of Process Control, 10(2):97–111, April 2000

George Stephanopoulos and Christine Ng. Perspectives on the synthesis of plant-wide control structures.Journal of Process Control, 10(2):97–111, April 2000

2000
[6]

Plantwide control: The search for the self-optimizing control structure

Sigurd Skogestad. Plantwide control: The search for the self-optimizing control structure. Journal of Process Control, 10(5):487–507, October 2000

2000
[7]

A review of methods for input/output selection

Marc van de Wal and Bram de Jager. A review of methods for input/output selection. Automatica, 37(4):487–510, April 2001

2001
[8]

Halvorsen, Sigurd Skogestad, John C

Ivar J. Halvorsen, Sigurd Skogestad, John C. Morud, and Vidar Alstad. Optimal selection of controlled variables.Industrial & Engineering Chemistry Research, 42(14):3273–3284, July 2003

2003
[9]

Null space method for selecting optimal measure- ment combinations as controlled variables.Industrial & Engineering Chemistry Research, 46(3):846–853, January 2007

Vidar Alstad and Sigurd Skogestad. Null space method for selecting optimal measure- ment combinations as controlled variables.Industrial & Engineering Chemistry Research, 46(3):846–853, January 2007

2007
[10]

Janardhanan

Vinay Kariwala, Yi Cao, and S. Janardhanan. Local self-optimizing control with aver- age loss minimization.Industrial & Engineering Chemistry Research, 47(4):1150–1158, February 2008

2008
[11]

Bidirectional branch and bound for controlled variable selection: Part i

Yi Cao and Vinay Kariwala. Bidirectional branch and bound for controlled variable selection: Part i. principles and minimum singular value criterion.Computers & Chemical Engineering, 32(10):2306–2319, October 2008

2008
[12]

Bidirectional branch and bound for controlled variable selection

Vinay Kariwala and Yi Cao. Bidirectional branch and bound for controlled variable selection. part ii: Exact local method for self-optimizing control.Computers and Chemical Engineering, 33(8):1402–1412, 2009. 45

2009
[13]

Bidirectional branch and bound for controlled variable selection part iii: Local average loss minimization.IEEE Transactions on Industrial Informatics, 6(1):54–61, February 2010

Vinay Kariwala and Yi Cao. Bidirectional branch and bound for controlled variable selection part iii: Local average loss minimization.IEEE Transactions on Industrial Informatics, 6(1):54–61, February 2010

2010
[14]

Approximating necessary conditions of optimality as controlled variables.Industrial and Engineering Chemistry Research, 2013

Lingjian Ye, Yi Cao, Yingdao Li, and Zhihuan Song. Approximating necessary conditions of optimality as controlled variables.Industrial and Engineering Chemistry Research, 2013

2013
[15]

Using process data for finding self-optimizing controlled variables*.IFAC Proceedings Volumes, 46(32):451–456, December 2013

Johannes Jäschke and Sigurd Skogestad. Using process data for finding self-optimizing controlled variables*.IFAC Proceedings Volumes, 46(32):451–456, December 2013

2013
[16]

Nco tracking and self-optimizing control in the context of real-time optimization.Journal of Process Control, 21(10):1407–1416, December 2011

Johannes Jäschke and Sigurd Skogestad. Nco tracking and self-optimizing control in the context of real-time optimization.Journal of Process Control, 21(10):1407–1416, December 2011

2011
[17]

Optimal controlled variables for polynomial systems.Journal of Process Control, 22(1):167–179, January 2012

Johannes Jäschke and Sigurd Skogestad. Optimal controlled variables for polynomial systems.Journal of Process Control, 22(1):167–179, January 2012

2012
[18]

Global approximation of self-optimizing con- trolled variables with average loss minimization.Industrial and Engineering Chemistry Research, 54(48):12040–12053, 2015

Lingjian Ye, Yi Cao, and Xiaofeng Yuan. Global approximation of self-optimizing con- trolled variables with average loss minimization.Industrial and Engineering Chemistry Research, 54(48):12040–12053, 2015

2015
[19]

Global self-optimizing control with active-set changes: A polynomial chaos approach.Computers & Chemical Engineering, 159:107662, March 2022

Lingjian Ye, Yi Cao, and Shuanghua Yang. Global self-optimizing control with active-set changes: A polynomial chaos approach.Computers & Chemical Engineering, 159:107662, March 2022

2022
[20]

Generalized global self-optimizing control for chemical processes part i

Lingjian Ye, Yi Cao, Yuchen He, Chenchen Zhou, Hongxin Su, Xinhui Tang, and Shuanghua Yang. Generalized global self-optimizing control for chemical processes part i. the existence of perfect controlled variables and numerical design methods.Industrial & Engineering Chemistry Research, 62(37):15051–15069, September 2023

2023
[21]

Self-optimizing control with active set changes

Henrik Manum and Sigurd Skogestad. Self-optimizing control with active set changes. Journal of Process Control, 22(5):873–883, June 2012

2012
[22]

Systematic design of active constraint switch- ing using classical advanced control structures.Industrial & Engineering Chemistry Re- search, 59(6):2229–2241, February 2020

Adriana Reyes-Lúa and Sigurd Skogestad. Systematic design of active constraint switch- ing using classical advanced control structures.Industrial & Engineering Chemistry Re- search, 59(6):2229–2241, February 2020. 46

2020
[23]

Systematic design of active constraint switching using selectors.Computers & Chemical Engineering, 143:107106, December 2020

Dinesh Krishnamoorthy and Sigurd Skogestad. Systematic design of active constraint switching using selectors.Computers & Chemical Engineering, 143:107106, December 2020

2020
[24]

Abunde Neba, Hoese M

F. Abunde Neba, Hoese M. Tornyeviadzi, Stein W. Østerhus, and Razak Seidu. Self- optimizing attainable regions of the anaerobic treatment process: Modeling performance targets under kinetic uncertainty.Water Research, 171:115377, March 2020

2020
[25]

Optimal control structure selection based on eco- nomics for continuous cross-flow grain dryer.Drying Technology, 41(10):1605–1619, July 2023

Agustín Bottari and Lautaro Braccia. Optimal control structure selection based on eco- nomics for continuous cross-flow grain dryer.Drying Technology, 41(10):1605–1619, July 2023

2023
[26]

Global self-optimizing control of a solid oxide fuel cell

Shengdong Fu and Lingjian Ye. Global self-optimizing control of a solid oxide fuel cell. In 2023 IEEE 12th Data Driven Control and Learning Systems Conference (DDCLS), pages 1865–1870, May 2023

2023
[27]

Efficiency maximization of a di- rect internal reforming solid oxide fuel cell in a two-layer self-optimizing control structure

Shengdong Fu, Lingjian Ye, Feifan Shen, and Yuchen He. Efficiency maximization of a di- rect internal reforming solid oxide fuel cell in a two-layer self-optimizing control structure. ACS Omega, 8(16):14558–14571, April 2023

2023
[28]

Two-stage anaerobic digestion process optimal control study based on extremum-seeking control and self-optimizing control

Hongxuan Li, Yang Tian, and Haoping Wang. Two-stage anaerobic digestion process optimal control study based on extremum-seeking control and self-optimizing control. In 2023 IEEE 12th Data Driven Control and Learning Systems Conference (DDCLS), pages 739–744, May 2023

2023
[29]

Integrateddesignand self-optimizing control of extractive distillation process with preconcentration.Chemical Engineering Science, 280:119074, October 2023

XiaodongZhang, ChengtianCui, JinshengSun, andXuemeiZhang. Integrateddesignand self-optimizing control of extractive distillation process with preconcentration.Chemical Engineering Science, 280:119074, October 2023

2023
[30]

Salsbury, and John M

Zhongfan Zhao, Yaoyu Li, Timothy I. Salsbury, and John M. House. Global self- optimizing control with data-driven optimal selection of controlled variables with ap- plication to chiller plant.Journal of Dynamic Systems, Measurement, and Control, 144(021008), November 2021

2021
[31]

Nikačević, Adrie E.M

Nikola M. Nikačević, Adrie E.M. Huesman, Paul M.J. Van den Hof, and Andrzej I. 47 Stankiewicz. Opportunities and challenges for process control in process intensification. Chemical Engineering and Processing: Process Intensification, 2012

2012
[32]

Biegler, Intan Hamdan, and John Wassick

Jun Shi, Lorenz T. Biegler, Intan Hamdan, and John Wassick. Optimization of grade transitions in polyethylene solution polymerization process under uncertainty.Computers & Chemical Engineering, 95:260–279, December 2016

2016
[33]

J. V. Kadam, W. Marquardt, B. Srinivasan, and D. Bonvin. Optimal grade transition in industrial polymerization processes via nco tracking.AIChE Journal, 53(3):627–639, March 2007

2007
[34]

Self-optimizingcontrol–asurvey.Annual Reviews in Control, 43:199–223, 2017

JohannesJäschke, YiCao, andVinayKariwala. Self-optimizingcontrol–asurvey.Annual Reviews in Control, 43:199–223, 2017

2017
[35]

L. T. Biegler and V. M. Zavala. Large-scale nonlinear programming using ipopt: An integrating framework for enterprise-wide dynamic optimization.Computers & Chemical Engineering, 33(3):575–582, March 2009

2009
[36]

Advanced control and on-line process optimization in multilayer struc- tures.Annual Reviews in Control, 32(1):71–85, April 2008

Piotr Tatjewski. Advanced control and on-line process optimization in multilayer struc- tures.Annual Reviews in Control, 32(1):71–85, April 2008

2008
[37]

Communications and Control Engineering

Lars Grüne and Jürgen Pannek.Nonlinear Model Predictive Control. Communications and Control Engineering. Springer International Publishing, Cham, 2017

2017
[38]

Christofides.Economic Model Predictive Control

Matthew Ellis, Jinfeng Liu, and Panagiotis D. Christofides.Economic Model Predictive Control. Advances in Industrial Control. Springer International Publishing, Cham, 2017

2017
[39]

Dufour, Y

P. Dufour, Y. Touré, D. Blanc, and P. Laurent. On nonlinear distributed parameter model predictive control strategy: On-line calculation time reduction and application to an experimental drying process.Computers & Chemical Engineering, 27(11):1533–1542, November 2003

2003
[40]

Stochastic model predictive control: An overview and perspectives for future research.IEEE Control Systems Magazine, 36(6):30–44, December 2016

Ali Mesbah. Stochastic model predictive control: An overview and perspectives for future research.IEEE Control Systems Magazine, 36(6):30–44, December 2016. 48

2016
[41]

Andersson, Heiko Brandt, Moritz Diehl, and Sebastian Engell

Sergio Lucia, Joel A.E. Andersson, Heiko Brandt, Moritz Diehl, and Sebastian Engell. Handling uncertainty in economic nonlinear model predictive control: A comparative case study.Journal of Process Control, 24(8):1247–1259, August 2014

2014
[42]

Dynamic considerations in the synthesis of self-optimizing control structures.AIChE Journal, 54(7):1830–1841, July 2008

Michael Baldea, Antonio Araujo, Sigurd Skogestad, and Prodromos Daoutidis. Dynamic considerations in the synthesis of self-optimizing control structures.AIChE Journal, 54(7):1830–1841, July 2008

2008
[43]

Grema and Yi Cao

Alhaji S. Grema and Yi Cao. Dynamic self-optimizing control for uncertain oil reservoir waterflooding processes.IEEE Transactions on Control Systems Technology, 28(6):2556– 2563, November 2020

2020
[44]

Selection of self-optimizing controlled variables for dynamic processes.IFAC Proceedings Volumes, 45(15):774–779, January 2012

Wuhua Hu, Jianfeng Mao, Gaoxi Xiao, and Vinay Kariwala. Selection of self-optimizing controlled variables for dynamic processes.IFAC Proceedings Volumes, 45(15):774–779, January 2012

2012
[45]

Dynamic self-optimizing control for unconstrained batch processes.Computers and Chemical Engineering, 117:451–468, 2018

Lingjian Ye and Sigurd Skogestad. Dynamic self-optimizing control for unconstrained batch processes.Computers and Chemical Engineering, 117:451–468, 2018

2018
[46]

Accounting for dynamics in self-optimizing control.Journal of Process Control, 76:15–26, April 2019

Jonatan Ralf Axel Klemets and Morten Hovd. Accounting for dynamics in self-optimizing control.Journal of Process Control, 76:15–26, April 2019

2019
[47]

Feedback control for optimal process operation.Journal of Process Control, 17(3):203–219, March 2007

Sebastian Engell. Feedback control for optimal process operation.Journal of Process Control, 17(3):203–219, March 2007

2007
[48]

Christofides

Matthew Ellis, Helen Durand, and Panagiotis D. Christofides. A tutorial review of eco- nomic model predictive control methods.Journal of Process Control, 24(8):1156–1178, August 2014

2014
[49]

Performance guaranteed mpc policy approximation via cost guided learning.IEEE Control Systems Letters, 8:346–351, 2024

Chenchen Zhou, Yi Cao, and Shuang-Hua Yang. Performance guaranteed mpc policy approximation via cost guided learning.IEEE Control Systems Letters, 8:346–351, 2024

2024
[50]

Bellman and Stuart E

Richard E. Bellman and Stuart E. Dreyfus.Applied Dynamic Programming. Princeton University Press, December 2015. 49

2015
[51]

Dontchev and R

Asen L. Dontchev and R. Tyrrell Rockafellar.Implicit Functions and Solution Mappings: A View from Variational Analysis. Springer Series in Operations Research and Financial Engineering. Springer New York, New York, NY, 2014

2014
[52]

A perspective on nonlinear model predictive control.Korean Journal of Chemical Engineering, 38(7):1317–1332, July 2021

Lorenz Theodor Biegler. A perspective on nonlinear model predictive control.Korean Journal of Chemical Engineering, 38(7):1317–1332, July 2021

2021
[53]

L. T. Biegler, X. Yang, and G. A.G. Fischer. Advances in sensitivity-based nonlinear model predictive control and dynamic real-time optimization.Journal of Process Control, 30:104–116, June 2015

2015
[54]

Bazaraa, Hanif D

Mokhtar S. Bazaraa, Hanif D. Sherali, and C. M. Shetty.Nonlinear Programming: Theory and Algorithms. John Wiley & Sons, June 2013

2013
[55]

Richard E. Bellman. Dynamic programming.Science, 153(3731):34–37, July 1966

1966
[56]

John C. Doyle. Guaranteed margins for lqg regulators.IEEE Transactions on Automatic Control, 1978

1978
[57]

Real-time optimization as a feedback control problem – a review.Computers & Chemical Engineering, 161:107723, May 2022

Dinesh Krishnamoorthy and Sigurd Skogestad. Real-time optimization as a feedback control problem – a review.Computers & Chemical Engineering, 161:107723, May 2022

2022
[58]

Bristow, M

D.A. Bristow, M. Tharayil, and A.G. Alleyne. A survey of iterative learning control. IEEE Control Systems Magazine, 26(3):96–114, June 2006

2006
[59]

An intelligent approach of controlled variable selection for constrained process self-optimizing control

Hongxin Su, Chenchen Zhou, Yi Cao, Shuang Hua Yang, and Zuzhen Ji. An intelligent approach of controlled variable selection for constrained process self-optimizing control. http://mc.manuscriptcentral.com/tssc, 10(1):65–72, 2022

2022
[60]

A heuristic for dynamic output predictive control design for uncertain nonlinear systems, February 2021

Mazen Alamir. A heuristic for dynamic output predictive control design for uncertain nonlinear systems, February 2021

2021
[61]

Joel A. E. Andersson, Joris Gillis, Greg Horn, James B. Rawlings, and Moritz Diehl. Casadi: A software framework for nonlinear optimization and optimal control.Mathe- matical Programming Computation, 11(1):1–36, March 2019. 50

2019
[62]

Ubrich, B

O. Ubrich, B. Srinivasan, F. Stoessel, and D. Bonvin. Optimization of a semi-batch reaction system under safety constraints. In1999 European Control Conference (ECC), pages 850–855, August 1999

1999
[63]

Srinivasan, S

B. Srinivasan, S. Palanki, and Dominique Bonvin. Dynamic optimization of batch pro- cesses: I. characterization of the nominal solution.Computers & Chemical Engineering, 27(1):1–26, January 2003

2003
[64]

D. Q. Mayne, E. C. Kerrigan, E. J. van Wyk, and P. Falugi. Tube-based robust non- linear model predictive control.International Journal of Robust and Nonlinear Control, 21(11):1341–1353, 2011

2011
[65]

Learning an approximate model predictive controller with guarantees.IEEE Control Systems Letters, 2018

Michael Hertneck, Johannes Kohler, Sebastian Trimpe, and Frank Allgower. Learning an approximate model predictive controller with guarantees.IEEE Control Systems Letters, 2018

2018
[66]

CRC Press, September 2019

Rein Luus.Iterative Dynamic Programming. CRC Press, September 2019

2019
[67]

Jorge J. Moré. The levenberg-marquardt algorithm: Implementation and theory. In G. A. Watson, editor,Numerical Analysis, pages 105–116, Berlin, Heidelberg, 1978. Springer

1978
[68]

Implicit behavioral cloning

PeteFlorence, CoreyLynch, AndyZeng, OscarA.Ramirez, AyzaanWahid, LauraDowns, Adrian Wong, Johnny Lee, Igor Mordatch, and Jonathan Tompson. Implicit behavioral cloning. InProceedings of the 5th Conference on Robot Learning, pages 158–168. PMLR, January 2022

2022
[69]

Using stochastic programming to train neural network approximation of nonlinear mpc laws.Automatica, 146:110665, December 2022

Yun Li, Kaixun Hua, and Yankai Cao. Using stochastic programming to train neural network approximation of nonlinear mpc laws.Automatica, 146:110665, December 2022

2022
[70]

January 2006

Yann Lecun, Sumit Chopra, Raia Hadsell, Marc’Aurelio Ranzato, and Fu-Jie Huang.A Tutorial on Energy-Based Learning. January 2006

2006
[71]

Laurent El Ghaoui, Fangda Gu, Bertrand Travacca, Armin Askari, and Alicia Y. Tsai. Implicit deep learning, August 2020

2020
[72]

An implicit function learning approach for parametric modal regression

Yangchen Pan, Ehsan Imani, Amir-massoud Farahmand, and Martha White. An implicit function learning approach for parametric modal regression. InAdvances in Neural In- 51 formation Processing Systems, volume 33, pages 11442–11452. Curran Associates, Inc., 2020. 52

2020