Proposal of a neural controller able to compensate neurological time delay for stance postural control

Abstract

In order to understand insights into the postural control mechanism, realization of biped stance motion in consideration of both muscles and neurological time delay (NTD) is important. In this paper, we proposed a neural controller composed of feed-forward (FF) and feedback (FB) control to keep a musculoskeletal model with 70 muscles standing stably when NTD was 100ms. FF control is modelled as a set of constant activations necessary to keep musculoskele-tal model standing in objective posture. FB control is modelled as a PD control based on muscular-tendon length and lengtening velocity. Parameter tuning was conducted to search feed-forward controls and feedback gains of PD control to keep musculoskeletal model standing. As a result, we successfully found the suitable parameters to keep musculoskeletal model standing when NTD was 100ms. We found FF control contribute more to compensate disturbances caused by NTD.