EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Domain-Knowledge-Guided Candidate Generation for Surrogate-Based Injection Molding Optimization

Yuxuan Xie, Xiaoyu Li, Haipeng Zou, Hao Dong, Ke Yao, Yongkuan Yang, Xiangsong Kong, Furong Gao

Processes · 13.11.2026 · 14 (22) · 3670

DOI: 10.3390/pr14223670 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Processes, online erschienen am 13.11.2026. Verfasst von Yuxuan Xie und 7 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: State Key Laboratory of Industrial Control Technology, Guangzhou Institute of Science and Technology, Fujian Provincial Department of Science and Technology.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Injection molding (IM) quality optimization involves coupled process variables, nonlinear quality responses, and a limited budget for physical trials. Numerical optimizers can use measured process data effectively, but they do not directly represent troubleshooting rules and other experience-based knowledge that is recorded in text. This study proposes large language model-based extended data-driven optimization (LLM-EDO), a domain-knowledge-guided framework for surrogate-based candidate generation. The framework converts process variables, quality curve discrepancies, feasible bounds, and retrieved expert cases into a structured request. A language model then proposes a bounded parameter vector, and a fixed long short-term memory surrogate evaluates the resulting quality curve. The framework is examined in a surrogate-in-the-loop environment using repeated and multi-start tests and is compared with Bayesian optimization, simultaneous perturbation stochastic approximation, and knowledge-informed gradient-based SPSA. The reported runs show reductions in curve distance for the tested initial configurations and competitive terminal objective values. They also show run-to-run variability, and iteration counts are reported as decision rounds rather than wall-clock measurements. The results support the feasibility of domain-knowledge-guided candidate generation for surrogate-based quality curve matching under the tested process configurations.

Förderung

  • State Key Laboratory of Industrial Control Technology
  • Guangzhou Institute of Science and Technology
  • Fujian Provincial Department of Science and Technology

Zitieren

APA

Xie, Y., Li, X., Zou, H., Dong, H., Yao, K., Yang, Y., Kong, X., & Gao, F. (2026). Domain-Knowledge-Guided Candidate Generation for Surrogate-Based Injection Molding Optimization. Processes, 14(22), 3670. https://doi.org/10.3390/pr14223670

Vancouver

Xie Y, Li X, Zou H, Dong H, Yao K, Yang Y, et al. Domain-Knowledge-Guided Candidate Generation for Surrogate-Based Injection Molding Optimization. Processes. 2026;14(22):3670. doi:10.3390/pr14223670

BibTeX
@article{Xie2026,
  author = {Xie, Yuxuan and Li, Xiaoyu and Zou, Haipeng and Dong, Hao and Yao, Ke and Yang, Yongkuan and Kong, Xiangsong and Gao, Furong},
  title = {Domain-Knowledge-Guided Candidate Generation for Surrogate-Based Injection Molding Optimization},
  journal = {Processes},
  year = {2026},
  volume = {14},
  number = {22},
  pages = {3670},
  doi = {10.3390/pr14223670}
}
Kostenlos starten →