A controller perspective on biological gait control: Reflexes and central pattern generators

摘要

Biological motion control paradigms can be analyzed from a control theory point of view and translated to control applications. This paper revises and updates important aspects about the neural control of biped gait based on the Central Pattern Generator (CPG) and provides a controller view, including possible artificial implementations of these controllers. A first aspect to consider is the biological sensor-actuation system because the particularities of this system: the mammalian muscle. Many current biped robots are based on symmetrical CPG. However, new evidence suggests that biological CPG is asymmetrical with rhythm generator controlling the flexor half in a kind of feedforward control and the extensor behaving in a feedback fashion controlled by external inputs such as ground contact. A theoretical model of the CPG connections with flexors and extensors is presented. (C) 2019 Published by Elsevier Ltd.