EUVIMEDObjevte EUVIMED
Open access · CC BY-NC-ND 4.0

Intelligent Task Planning for Reconfigurable Mechanisms Using Reinforcement Learning Based Structural Switching Strategy

Daniel J. Carter, Xiaoling Wu, Emily R. Davies, Thomas H. Bennett, Olivia M. Clarke

Frontiers in Artificial Intelligence Research · 2025-09-19 · 2 (3) · 345-351

DOI: 10.71465/fair361 ↗

At a glance

Journal article in Frontiers in Artificial Intelligence Research, published online on 2025-09-19. Written by Daniel J. Carter and 4 co-authors. The full text is freely available (open access, CC BY-NC-ND 4.0).

AbstraktBibliografické údajeCiteRelated studies

Abstract (original language)

Reconfigurable mechanisms can achieve multi-task execution through structural switching, but determining the optimal reconfiguration strategy in complex environments remains a major challenge. This study proposes an intelligent task planning method based on reinforcement learning. The structural switching process is modeled as a Markov decision process, where the action space corresponds to topology changes and the reward function jointly considers task completion rate and energy consumption. A deep Q-network is employed to train the optimal switching strategy. Experiments conducted in 15 task environments demonstrate that the proposed method achieves a 35% improvement in task completion rate and an 18% reduction in average energy consumption compared with baseline search algorithms. Moreover, after training, the decision-making speed on the simulation platform is approximately 10 times faster than that of traditional search methods. These results confirm that reinforcement learning can significantly enhance both efficiency and adaptability in reconfigurable mechanisms, providing an effective pathway for intelligent control of reconfigurable robots and adaptive mechanical systems.

Cite

APA

Carter, D. J., Wu, X., Davies, E. R., Bennett, T. H., & Clarke, O. M. (2025). Intelligent Task Planning for Reconfigurable Mechanisms Using Reinforcement Learning Based Structural Switching Strategy. Frontiers in Artificial Intelligence Research, 2(3), 345-351. https://doi.org/10.71465/fair361

Vancouver

Carter DJ, Wu X, Davies ER, Bennett TH, Clarke OM. Intelligent Task Planning for Reconfigurable Mechanisms Using Reinforcement Learning Based Structural Switching Strategy. Frontiers in Artificial Intelligence Research. 2025;2(3):345-351. doi:10.71465/fair361

BibTeX
@article{Carter2025,
  author = {Carter, Daniel J. and Wu, Xiaoling and Davies, Emily R. and Bennett, Thomas H. and Clarke, Olivia M.},
  title = {Intelligent Task Planning for Reconfigurable Mechanisms Using Reinforcement Learning Based Structural Switching Strategy},
  journal = {Frontiers in Artificial Intelligence Research},
  year = {2025},
  volume = {2},
  number = {3},
  pages = {345--351},
  doi = {10.71465/fair361}
}
Start for free →