Arrow Research search

Author name cluster

Dongkun Wang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

2 papers
1 author row

Possible papers

2

EAAI Journal 2025 Journal Article

A novel contrastive learning framework for multi-parameter optimization in 3D printing

  • Jieyang Peng
  • Simon Kreuzwieser
  • Dongkun Wang
  • Andreas Kimmig
  • Zhi Fan
  • Jianing Li
  • Jivka Ovtcharova

Additive manufacturing (3D printing) revolutionizes prototyping and production through unparalleled material efficiency. However, part quality remains highly sensitive to parameter variations, where subtle deviations induce defects, material waste, and process instability. Traditional quality control methods – reliant on rule-based heuristics or manual inspections – fail to address complex multi-parameter interactions and fine-grained anomalies, limiting industrial scalability. This article focuses on an important issue in 3D printing: How can we develop a robust, automated framework to simultaneously detect and classify subtle multi-parameter anomalies in 3D printing, overcoming the limitations of manual and single-defect-focused approaches? We propose a supervised contrastive learning framework integrating Vision Transformers (ViT) to learn discriminative feature representations for multi-parameter optimization. By maximizing intra-class similarity and inter-class separation, our model captures nuanced variations across printing scenarios. The ViT architecture processes real-time printing images, while contrastive loss ensures compact feature clusters for “Low”, “Optimal”, and “High” parameter classes. Experimental evaluations on open-source datasets demonstrate our framework achieves 8. 45% accuracy, outperforming conventional CNNs by 10. 11%. Real-world validation shows robust performance across critical parameters: flow rate (86. 5% accuracy in nominal ranges), feed rate (87% accuracy), and extrusion temperature (90% accuracy at optimal settings). The ViT’s self-attention mechanism enables precise detection of localized anomalies, such as under-extrusion and layer misalignment.

AAAI Conference 2023 Conference Paper

Dish-TS: A General Paradigm for Alleviating Distribution Shift in Time Series Forecasting

  • Wei Fan
  • Pengyang Wang
  • Dongkun Wang
  • Dongjie Wang
  • Yuanchun Zhou
  • Yanjie Fu

The distribution shift in Time Series Forecasting (TSF), indicating series distribution changes over time, largely hinders the performance of TSF models. Existing works towards distribution shift in time series are mostly limited in the quantification of distribution and, more importantly, overlook the potential shift between lookback and horizon windows. To address above challenges, we systematically summarize the distribution shift in TSF into two categories. Regarding lookback windows as input-space and horizon windows as output-space, there exist (i) intra-space shift, that the distribution within the input-space keeps shifted over time, and (ii) inter-space shift, that the distribution is shifted between input-space and output-space. Then we introduce, Dish-TS, a general neural paradigm for alleviating distribution shift in TSF. Specifically, for better distribution estimation, we propose the coefficient net (Conet), which can be any neural architectures, to map input sequences into learnable distribution coefficients. To relieve intra-space and inter-space shift, we organize Dish-TS as a Dual-Conet framework to separately learn the distribution of input- and output-space, which naturally captures the distribution difference of two spaces. In addition, we introduce a more effective training strategy for intractable Conet learning. Finally, we conduct extensive experiments on several datasets coupled with different state-of-the-art forecasting models. Experimental results show Dish-TS consistently boosts them with a more than 20% average improvement. Code is available at https://github.com/weifantt/Dish-TS.

v2026.09.13