EUVIMEDEUVIMED entdecken
Crossref

Energy-Aware Trajectory Optimization Using Deep Reinforcement Learning in Humanoid Robot Through Via-Point Learning

James Sorokhaibam, Adersh Maruvattu, Ashish Dutta

Journal of Mechanisms and Robotics · 01.08.2026 · 18 (8)

DOI: 10.1115/1.4072208 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Mechanisms and Robotics, online erschienen am 10.07.2026. Verfasst von James Sorokhaibam und 2 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Learning whole-body motion for a 19-degrees-of-freedom (DoFs) humanoid robot using deep reinforcement learning (DRL) is challenging due to high kinematic redundancy, large joint-level search spaces, and the need to maintain dynamic balance. This article proposes a structured DRL framework that integrates task-space via-point trajectory learning with a pseudo-inverse of Jacobian matrix for redundancy resolution (PJRR) method. Instead of learning full joint trajectories, a soft actor-critic (SAC) agent predicts a compact set of task-space via-points, which are converted into C2-continuous motions using composite cubic splines. Kinematic redundancy is resolved analytically through the PJRR, which operates separately from the learned policy and enforces a three-level task hierarchy: end-effector trajectory subtask, hip trajectory subtask, and posture regulation subtask. The framework is validated on the KONDO KHR-3HV humanoid platform (1.3 kg, 19-DoFs) across four pick-and-place configurations, evaluated over N=24 independent trials per motion. Energy-aware trajectories are generated with no zero moment point (ZMP) violations in simulation. A time-incentive reward formulation characterizes the energy-time trade-off, and a time-conditioned policy enables speed generalization at inference without retraining. Preliminary transferability to a scaled 23-DoFs platform is demonstrated in simulation.

Zitieren

APA

Sorokhaibam, J., Maruvattu, A., & Dutta, A. (2026). Energy-Aware Trajectory Optimization Using Deep Reinforcement Learning in Humanoid Robot Through Via-Point Learning. Journal of Mechanisms and Robotics, 18(8). https://doi.org/10.1115/1.4072208

Vancouver

Sorokhaibam J, Maruvattu A, Dutta A. Energy-Aware Trajectory Optimization Using Deep Reinforcement Learning in Humanoid Robot Through Via-Point Learning. Journal of Mechanisms and Robotics. 2026;18(8). doi:10.1115/1.4072208

BibTeX
@article{Sorokhaibam2026,
  author = {Sorokhaibam, James and Maruvattu, Adersh and Dutta, Ashish},
  title = {Energy-Aware Trajectory Optimization Using Deep Reinforcement Learning in Humanoid Robot Through Via-Point Learning},
  journal = {Journal of Mechanisms and Robotics},
  year = {2026},
  volume = {18},
  number = {8},
  doi = {10.1115/1.4072208}
}
Kostenlos starten →