arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2007.06669eess.SPcs.LGcs.ROcs.SYeess.SY
收藏

基于功能模拟的肌肉骨骼控制强化学习

Reinforcement Learning of Musculoskeletal Control from Functional Simulations

ETH Zurich(苏黎世联邦理工学院)

Emanuel Joos, Fabien Péan, Orcun Goksel

摘要:

为了诊断、计划和治疗肌肉骨骼疾病,理解和再现复杂运动的肌肉招募是必不可少的。由于运动的肌肉激活通常是高度冗余、非线性和时间依赖的,机器学习可以为特定于解剖结构的肌肉骨骼模拟提供建模和控制的解决方案。复杂的生物力学模拟通常需要专门的计算环境,数值复杂且缓慢,阻碍了它们与典型的深度学习框架的集成。在这项工作中,基于深度强化学习(DRL)的逆动力学控制器被训练来控制人体肩部生物力学模型的肌肉激活。在可推广的端到端方式中,肌肉激活是在给定当前和期望的位置-速度对的情况下学习的。引入了用于轨迹控制的定制奖励函数,从而能够直接扩展到额外的肌肉和更高的自由度。使用生物力学模型,使用正在训练的DRL的进化神经模型同时在集群上模拟多个事件。结果提出了一个单轴运动控制肩外展的任务,随机生成的角轨迹。

英文摘要:

To diagnose, plan, and treat musculoskeletal pathologies, understanding and reproducing muscle recruitment for complex movements is essential. With muscle activations for movements often being highly redundant, nonlinear, and time dependent, machine learning can provide a solution for their modeling and control for anatomy-specific musculoskeletal simulations. Sophisticated biomechanical simulations often require specialized computational environments, being numerically complex and slow, hindering their integration with typical deep learning frameworks. In this work, a deep reinforcement learning (DRL) based inverse dynamics controller is trained to control muscle activations of a biomechanical model of the human shoulder. In a generalizable end-to-end fashion, muscle activations are learned given current and desired position-velocity pairs. A customized reward functions for trajectory control is introduced, enabling straightforward extension to additional muscles and higher degrees of freedom. Using the biomechanical model, multiple episodes are simulated on a cluster simultaneously using the evolving neural models of the DRL being trained. Results are presented for a single-axis motion control of shoulder abduction for the task of following randomly generated angular trajectories.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑