Automatic differentiation rules for Tsoukalas–Mitsos convex relaxations in global process optimization
Bibliographic record
Abstract
With increasing digitalization and vertical integration of chemical process systems, nonconvex optimization problems often emerge in chemical engineering applications, yet require specialized optimization techniques. Typical global optimization methods proceed by progressively refining bounds on the unknown optimal value, by strategically employing convex relaxations. This article constructs a general closed-form expression for the convex subdifferentials of recent “multivariate McCormick” convex relaxations of nontrivial composite functions, by solving a previous duality formulation in all cases using nonsmooth Karush–Kuhn–Tucker conditions. Based on this subdifferential expression, new automatic differentiation rules are developed to compute gradients and subgradients for multivariate McCormick relaxations, to ultimately generate useful bounds in global optimization. Unlike established differentiation techniques for these relaxations, our new rules are expressed in closed form, do not require solving separate dual optimization problems, are efficiently carried out, and are compatible with the reverse/adjoint mode of algorithmic differentiation. Our formulations become more straightforward when the relevant functions are either smooth or piecewise smooth. • Closed-form subdifferentials are obtained for Tsoukalas–Mitsos convex relaxations. • Forward-mode and reverse-mode automatic differentiation rules are presented. • Straightforward to implement; implemented in MATLAB for illustration. • An effective tool for lower-bounding in global optimization of process models. • Permits including new elemental functions in the Tsoukalas–Mitsos relaxation library.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".