Human arms can produce stable and variable compliant joint motions to achieve tasks in various spatial tasks and temporal scales. To emulate such motions we propose an adaptive motor controller (AMC) allowing for spatial-temporal adaptation of human-like motor control. The AMC is a biomimetic controller consisting of online force and impedance (i.e., stiffness and damping) adaptation to different tasks and unknown arm dynamics. As a result, the AMC can produce more accurate and stable human-like reaching and tracking behaviors, compared to conventional controllers. Moreover, the reproduced spatial-temporal adaptation is comparable to that found in the experiments of human motor control. The proposed AMC may pave a novel and simple way forward to understanding and solving inverse dynamics and variable impedance control in robotics and biomechanics.
|Titel||2018 IEEE International Conference on Robotics and Biomimetics, ROBIO 2018|
|Status||Udgivet - mar. 2019|
|Begivenhed||2018 IEEE International Conference on Robotics and Biomimetics, ROBIO 2018 - Kuala Lumpur, Malaysia|
Varighed: 12. dec. 2018 → 15. dec. 2018
|Konference||2018 IEEE International Conference on Robotics and Biomimetics, ROBIO 2018|
|Periode||12/12/2018 → 15/12/2018|
|Sponsor||et al., Guangdong University of Technology, Guangzhou University, IEEE, IEEE Robotics and Automation Society (RAS), Shenzhen Han's Robot Co., Ltd.|