EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

A Novel Deep Reinforcement Learning Approach for Dynamic Proportional‐Integral Control in Scanning Probe Microscopy

Ziwei Wei, Shuming Wei, Qibin Zeng, Wanheng Lu, Huajun Liu, Kaiyang Zeng iD

Small · 2025-09 · 21 (37)

DOI: 10.1002/smll.202503921 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Small, online erschienen am 30.07.2025. Verfasst von Ziwei Wei und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore, 117576, Singapore. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National University of Singapore, Academic Research Funds, NUS Research Scholarship.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Scanning Probe Microscopy (SPM) measurements often encounter challenges due to the nonlinear and time‐varying behavior of materials, as well as abrupt changes in structure, properties, or topography. These issues can lead to system instability and imaging artifacts. Traditional proportional‐integral (P‐I) controllers with fixed parameters struggle to adapt to such conditions. To address this limitation, this study introduces the Parallel Integrated Control and Training System (PICTS), which leverages deep reinforcement learning (DRL) to dynamically adjust control strategies in real time. This approach stabilizes probe–sample interactions, even for samples with sharp edges, soft multiphase materials, or complex topographies. Experimental results demonstrate that the DRL‐based controller reduces deflection errors by 26% – 90% compared to commercial fixed‐parameter controllers, producing more stable images with fewer artifacts. Statistical analyses confirm improved precision, with error values concentrated near zero. Additionally, the system employs a field‐programmable gate array (FPGA) for critical tasks, ensuring efficient operation without the need for high‐performance computing. By integrating advanced machine learning, this work enhances imaging quality and stability in demanding environments, paving the way for further innovations in SPM.

Schlagwörter

Stichwörter der Autor:innen

adaptive controldeep reinforcement learningfield‐programmable gate arrayparallel integrated control and training systemproportional‐integral controlscanning probe microscopysoft actor‐critic

Autor:innen und Institutionen

  • Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore, 117576, Singapore.
  • Institute of Materials Research and Engineering, Agency for Science, Technology and Research (A*STAR), 2 Fusionopolis Way, Innovis, #08-03, Singapore, 138634, Singapore.
  • Department of Electrical and Computer Engineering, National University of Singapore, 4 Engineering Drive 3, Singapore, 117583, Singapore.
  • NUS Research Institute (NUSRI), No. 377 Linquan Street, Suzhou Industrial Park, Suzhou, Jiangsu, 215123, China.

ORCID-Profile: Kaiyang Zeng

Förderung

  • National University of Singapore
  • Academic Research Funds
  • NUS Research Scholarship

Zitieren

APA

Wei, Z., Wei, S., Zeng, Q., Lu, W., Liu, H., & Zeng, K. (2025). A Novel Deep Reinforcement Learning Approach for Dynamic Proportional‐Integral Control in Scanning Probe Microscopy. Small, 21(37). https://doi.org/10.1002/smll.202503921

Vancouver

Wei Z, Wei S, Zeng Q, Lu W, Liu H, Zeng K. A Novel Deep Reinforcement Learning Approach for Dynamic Proportional‐Integral Control in Scanning Probe Microscopy. Small. 2025;21(37). doi:10.1002/smll.202503921

BibTeX
@article{Wei2025,
  author = {Wei, Ziwei and Wei, Shuming and Zeng, Qibin and Lu, Wanheng and Liu, Huajun and Zeng, Kaiyang},
  title = {A Novel Deep Reinforcement Learning Approach for Dynamic Proportional‐Integral Control in Scanning Probe Microscopy},
  journal = {Small},
  year = {2025},
  volume = {21},
  number = {37},
  doi = {10.1002/smll.202503921}
}
Kostenlos starten →