PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
May 1, 2026Applied Sciences0 citationsOpen Access

Small-Data Neural Computing Outperforms RSM: Low-Cost Smart Optimization in Injection Molding

View Full Paper
MYMing-Lang YehWPWen PeiHHHan-Ching Huang

Key Points

  • The aim is to enhance optimization in injection molding through small data neural computing techniques amidst data scarcity.
  • Utilized a domain-knowledge guided data augmentation framework combining Taguchi experimental data and Moldex3D simulations.
  • Employed a back-propagation neural network with L2 regularization for small-sample learning.
  • Conducted evaluation using nested group-based 5-fold cross-validation to avoid data leakage.
  • BPNN achieved a testing mean squared error of 0.001 (±0.0002) and a testing R² of 0.95, outperforming RSM and RF.
  • The shrinkage rate was minimized to 3.079% using PSO, validated with a 0.19% relative error from simulations.

Abstract

In smart manufacturing, the injection molding industry faces a “data scarce environment” due to prohibitive physical trial costs. Processing recycled polypropylene (rPP) exacerbates this challenge, as traditional response surface methodology (RSM) fails to capture complex non-linear rheological behaviors induced by material variability. This study proposes a “domain-knowledge guided data augmentation framework,” integrating Taguchi experimental data (L25) with Moldex3D digital twin simulations to construct a 300-sample hybrid dataset. A back-propagation neural network (BPNN) with L2 regularization was employed for small-sample learning, providing a continuous differentiable physical mapping. To rigorously prevent neighborhood data leakage, the model was evaluated via a strict nested group-based 5-fold cross-validation. Particle swarm optimization (PSO) was coupled to overcome the local minima of gradient descent. Comparative analysis demonstrates that BPNN significantly outperforms both traditional RSM and a newly introduced Random Forest (RF) baseline, achieving a testing mean squared error (MSE) of 0.001 (±0.0002) and a testing R2 of 0.95. PSO minimized the shrinkage rate to 3.079%, validated via Moldex3D digital twin simulation with a 0.19% relative error. Synergizing virtual–physical integration with robust neural computing enables superior process control precision in small-data regimes, offering small and medium-sized enterprises (SMEs) a cost-effective pathway for smart optimization.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Yeh et al. (2026) studied this question.

synapsesocial.com/papers/69f443e8967e944ac556700dhttps://doi.org/10.3390/app16094288
Ask AI
Helpful
Bookmark
Share
View Full Paper