Data-driven multi-objective optimization of flow field header design for PEM fuel cells
Bibliographic record
Abstract
Optimizing reactant distribution in flow field plates is critical for proton exchange membrane (PEM) fuel cell performance. In this study, cathode flow field header designs are explored using a hybrid framework that integrates computational fluid dynamics (CFD), artificial neural networks (ANN), and the non-dominated sorting genetic algorithm II (NSGA-II). Twenty-seven CFD simulations, generated by varying header design, including its porosity (the ratio of fluid area with obstacles to that without obstacles), header size, and obstacle size, provide the dataset for training a multi-input multi-output surrogate ANN model, whereas the number and dimensions of flow channels in the active area are fixed to isolate header effects, with identical inlet and outlet headers to reduce design complexity, maintain symmetry, and ensure consistence and comparability. The trained model achieves high accuracy (R 2 = 0.999) and enables rapid evaluation of design alternatives. Multi-objective optimization through NSGA-II yields a Pareto front balancing flow uniformity and pressure drop. The optimized design achieves flow uniformity >92 % with a pressure drop of ∼1900 Pa, closely matching the CFD simulation outcomes. This integrated, data-driven approach lowers computational cost, accelerates header design exploration, and offers a practical pathway for advancing PEM fuel cell technology toward commercialization. • CFD-ANN-NSGA-II framework optimizes PEM fuel cell cathode header design. • Dataset of 27 CFD cases trains a high-accuracy ANN surrogate model (R 2 = 0.999). • NSGA-II optimization balances flow uniformity (>92 %) and pressure drop (∼1900 Pa). • Knee-point solution shows <1 % error compared to CFD validation results. • Developed framework cuts computational cost and aids scalable PEM fuel cell design.
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 | 0.001 | 0.001 |
| 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 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".