EUVIMEDEUVIMED entdecken
Crossref

A Nonlinear MPC-Net Optimization Framework for Wheeled Humanoid Robots With Whole Body Dynamics

Guoxin Li, Xingjian Liu, Xiaoyu Wu, Qirong Tang, Hao Zhang, Zhijun Li, Peng Shi

IEEE Transactions on Automation Science and Engineering · 2026 · 23 · 3410-3422

DOI: 10.1109/tase.2025.3552208 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Automation Science and Engineering, 2026. Verfasst von Guoxin Li und 6 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 49 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Dreams Foundation of Jianghuai Advance Technology Center, Open Project of Fund of Robot Technology Used for Special Environment Key Laboratory of Sichuan Province, Southwest University of Science and Project of Techno….

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The task performance of mobile manipulators can be significantly enhanced by whole-body control and optimization in complex scenarios. Due to the nonlinear properties of whole-body dynamics and parameter uncertainty, modeling accurate system dynamics is essential in addition to designing an effective control strategy. However, traditional control methods have high computational costs and fail to deal with the parameter errors caused by model linearization. To address these issues, we propose a model predictive control (MPC)-Net, a learning-based approach that facilitates rapid online optimization by combining deep learning with multiple MPCs. Firstly, we develop a parameter identification algorithm based on a deep neural model to estimate the unknown dynamics parameters. Although the control performance of the MPC approach positively correlated with the prediction horizon, a long horizon would result in additional computational costs. Thus, MPC-Net is constructed by combining multiple sub-MPC issues, and the nonlinear coefficients are obtained by using a deep neural network-based optimization framework. Furthermore, MPC-Net generates the solution by combining the outputs of multiple sub-MPC problems using the nonlinear transformation of learned coefficients. Experiments are conducted on a mobile manipulator, which demonstrates the proposed MPC-Net-based optimization control offers fast efficient computation and low tracking error performance. Note to Practitioners—In the evolving landscape of industrial automation, enhancing task performance of mobile manipulators is crucial. Recent advancements in whole-body control and optimization have yielded promising results, particularly in complex scenarios where traditional methods often fall short due to high computational costs and difficulties in accurately modeling system dynamics. Our proposed solution, MPC-Net, addresses these issues by leveraging model predictive control combined with deep learning techniques. By integrating multiple sub-MPC solutions, MPC-Net utilizes nonlinear coefficients optimized through deep learning, ensuring efficient computations associated with long prediction horizons and reduced tracking errors. Experimental results on mobile manipulators demonstrate that MPC-Net significantly enhances control performance by learning rules from data, making it a practical choice for industrial applications. In the future, the integration of MPC-Net with more Artificial Intelligence Generated Content (AIGC) methods will achieve faster, more reliable task execution in real operations by creating realistic simulations for training and refining control algorithms, ultimately driving industrial automation applications forward.

Förderung

  • National Natural Science Foundation of China
  • Dreams Foundation of Jianghuai Advance Technology Center
  • Open Project of Fund of Robot Technology Used for Special Environment Key Laboratory of Sichuan Province, Southwest University of Science and Project of Techno…
  • Natural Science Foundation of Anhui Province

Zitieren

APA

Li, G., Liu, X., Wu, X., Tang, Q., Zhang, H., Li, Z., & Shi, P. (2026). A Nonlinear MPC-Net Optimization Framework for Wheeled Humanoid Robots With Whole Body Dynamics. IEEE Transactions on Automation Science and Engineering, 23, 3410-3422. https://doi.org/10.1109/tase.2025.3552208

Vancouver

Li G, Liu X, Wu X, Tang Q, Zhang H, Li Z, et al. A Nonlinear MPC-Net Optimization Framework for Wheeled Humanoid Robots With Whole Body Dynamics. IEEE Transactions on Automation Science and Engineering. 2026;23:3410-3422. doi:10.1109/tase.2025.3552208

BibTeX
@article{Li2026,
  author = {Li, Guoxin and Liu, Xingjian and Wu, Xiaoyu and Tang, Qirong and Zhang, Hao and Li, Zhijun and Shi, Peng},
  title = {A Nonlinear MPC-Net Optimization Framework for Wheeled Humanoid Robots With Whole Body Dynamics},
  journal = {IEEE Transactions on Automation Science and Engineering},
  year = {2026},
  volume = {23},
  pages = {3410--3422},
  doi = {10.1109/tase.2025.3552208}
}
Kostenlos starten →