Loading [a11y]/accessibility-menu.js
Proposal of Feedforward Trajectory Control with Iterative Learning for a Musculoskeletal System | IEEE Conference Publication | IEEE Xplore

Proposal of Feedforward Trajectory Control with Iterative Learning for a Musculoskeletal System

Publisher: IEEE

Abstract:

As one of the posture controls in the musculoskeletal system, the musculoskeletal potential method has been studied. The musculoskeletal potential method utilizes the pro...View more

Abstract:

As one of the posture controls in the musculoskeletal system, the musculoskeletal potential method has been studied. The musculoskeletal potential method utilizes the property of a potential generated by the internal force among muscles. By the step input of muscular tension balancing at the desired posture, posture control can be achieved without any sensory feedback or complicated real-time calculations. The previous study extended the feedforward point-to-point position control method to feedforward trajectory control. In this study, trajectory control can be achieved by optimizing the muscular internal force that forms the potential field to realize the target trajectory. However, the previous trajectory control had a limitation to be impossible to create an arbitrary potential field even if the muscular internal force was changed. Consequently, the previous trajectory control was suitable for only quite limited desired trajectories. To overcome such a difficulty, this paper proposes a new method to achieve feedforward trajectory control through iterative trials. In the proposed method, the controller learns the initial joint angular velocities and joint viscosity coefficients by trials to achieve the feedforward trajectory control generated by the step input of muscular internal force. The effectiveness of the proposed method is verified through numerical simulations by using a musculoskeletal system with two joints and six muscles.
Date of Conference: 28 February 2024 - 01 March 2024
Date Added to IEEE Xplore: 26 April 2024
ISBN Information:

ISSN Information:

Publisher: IEEE
Conference Location: Kyoto, Japan

References

References is not available for this document.