EUVIMEDEUVIMED entdecken
Crossref

A Novel Neural-Network-Based MPC Framework for Whole-Body Motion Optimization and Control of Redundant Humanoid Robots

Jie Wang, Yiming Jiang, Yu Li, Hui Zhang, Yaonan Wang

IEEE Transactions on Automation Science and Engineering · 2026 · 23 · 7920-7931

DOI: 10.1109/tase.2026.3679503 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Automation Science and Engineering, 2026. Verfasst von Jie Wang und 4 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Gefördert von: National Key Research and Development Program of China, National Natural Science Foundation of China, Changsha Outstanding Innovative Youth Training Program.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The rapid advancement of humanoid robotics has highlighted the challenge of coordinating high-redundancy whole-body motion and posture control under multiple safety constraints. The difficulty lies particularly in dual-arm task planning and the design of high-precision, real-time control strategies. To address these issues, this paper proposes a novel neural-network (NN) based real-time model predictive control (MPC) framework for humanoid robots. The framework innovatively integrates discrete recurrent neural networks (DRNN) with MPC, thereby extending their combined advantages to high-redundancy humanoid motion control. In addition, a primal-dual neural network (PDNN) solver is employed to compute multi-constrained kinematic MPC in a single iteration, avoiding the repeated iterations required in conventional MPC and significantly enhancing real-time performance. The proposed approach is rigorously validated through theoretical derivation and extensive experiments, including both numerical simulations and real-world trials. After theoretical verification in simulation, the NN-based MPC framework is deployed on a self-developed humanoid robotic platform. Experimental results confirm that the NN-based MPC method achieves effective and highly accurate whole-body task planning and real-time control, demonstrating its potential as a reliable solution for advanced humanoid robot control. Note to Practitioners—This study presents a novel NN-based MPC framework for high-redundancy humanoid robots, aiming to improve motion control accuracy and real-time performance in complex tasks. Traditional MPC struggles to meet real-time requirements in high-dimensional, multi-constraint scenarios. To overcome this, we integrate the novel NN-based MPC with a PDNN solver, enabling multi-constrained motion control in a single iteration, which enhances efficiency. The key innovation of this research lies in combining MPC with neural networks to accurately track the position and orientation of the robot’s end-effectors. This method effectively handles multiple constraints while maintaining optimization performance, avoiding the iterative process common in traditional MPC, and achieving real-time optimization. The proposed NN-based MPC framework has been successfully deployed and validated on a self-developed humanoid robotic platform. The experimental results demonstrate that the proposed NN-based MPC, by incorporating collaborative metrics, effectively accomplishes cooperative transport tasks. This capability positions it as a practical choice for industrial applications.

Förderung

  • National Key Research and Development Program of China
  • National Natural Science Foundation of China
  • Changsha Outstanding Innovative Youth Training Program
  • Science and Technology Innovation Program of Hunan Province
  • Yuelushan Center for Industrial Innovation

Zitieren

APA

Wang, J., Jiang, Y., Li, Y., Zhang, H., & Wang, Y. (2026). A Novel Neural-Network-Based MPC Framework for Whole-Body Motion Optimization and Control of Redundant Humanoid Robots. IEEE Transactions on Automation Science and Engineering, 23, 7920-7931. https://doi.org/10.1109/tase.2026.3679503

Vancouver

Wang J, Jiang Y, Li Y, Zhang H, Wang Y. A Novel Neural-Network-Based MPC Framework for Whole-Body Motion Optimization and Control of Redundant Humanoid Robots. IEEE Transactions on Automation Science and Engineering. 2026;23:7920-7931. doi:10.1109/tase.2026.3679503

BibTeX
@article{Wang2026,
  author = {Wang, Jie and Jiang, Yiming and Li, Yu and Zhang, Hui and Wang, Yaonan},
  title = {A Novel Neural-Network-Based MPC Framework for Whole-Body Motion Optimization and Control of Redundant Humanoid Robots},
  journal = {IEEE Transactions on Automation Science and Engineering},
  year = {2026},
  volume = {23},
  pages = {7920--7931},
  doi = {10.1109/tase.2026.3679503}
}
Kostenlos starten →