Energy-Efficient Prediction in Textile Manufacturing: Enhancing Accuracy and Data Efficiency With Ensemble Deep Transfer Learning

TL;DR

This work addresses the challenge of energy-efficient textile production under data scarcity by introducing Ensemble Deep Transfer Learning (EDTL), which pre-trains deep models on data-rich production lines and adapts them to data-poor lines using an ensemble of base learners and a feature-alignment layer. The method combines network-based deep transfer learning with stacking regression (SVR) to produce accurate, multi-target predictions for electricity consumption and fabric quality indicators, even with limited target-domain data. Key contributions include applying DTL to textile manufacturing for cross-line adaptation, mitigating negative transfer via ensemble diversity, and achieving robust performance under anomalous data while enabling multi-target predictions. Experimental results on real factory data show EDTL improves prediction accuracy by $5.66\%$ and robustness by $3.96\%$ relative to conventional DNNs, with pronounced gains in data-scarce regimes ($20\%$–$40\%$ data). The approach supports scalable, data-efficient smart production and can be extended with adaptive anomaly detection and uncertainty quantification for broader industrial impact.

Abstract

Traditional textile factories consume substantial energy, making energy-efficient production optimization crucial for sustainability and cost reduction. Meanwhile, deep neural networks (DNNs), which are effective for factory output prediction and operational optimization, require extensive historical data, posing challenges due to high sensor deployment and data collection costs. To address this, we propose Ensemble Deep Transfer Learning (EDTL), a novel framework that enhances prediction accuracy and data efficiency by integrating transfer learning with an ensemble strategy and a feature alignment layer. EDTL pretrains DNN models on data-rich production lines (source domain) and adapts them to data-limited lines (target domain), reducing dependency on large datasets. Experiments on real-world textile factory datasets show that EDTL improves prediction accuracy by 5.66% and enhances model robustness by 3.96% compared to conventional DNNs, particularly in data-limited scenarios (20%-40% data availability). This research contributes to energy-efficient textile manufacturing by enabling accurate predictions with fewer data requirements, providing a scalable and cost-effective solution for smart production systems.

Figures (11)

• Figure 1: Stenter Setting Machine (SSM).
• Figure 2: Flowchart of generating control parameters.
• Figure 3: Proposed workflow for constructing prediction models.
• Figure 4: Proposed EDTL.
• Figure 5: Data collection platform in textile factories.