Abstract
In order to accurately obtain the motion information of the lower limbs of the human body and improve the control accuracy of the wearable lower limb exoskeleton. In this paper, on the basis of analyzing the characteristics of human lower limb movement joints, a set of human lower limb movement perception system based on multi-sensor information fusion is designed. The modules communicate quickly via the CAN bus. And study the use of Gated Recurrent Unit (GRU) to make short-term prediction of sensor signals to solve the problem of exoskeleton control hysteresis caused by sensor signal transmission delay. Finally, the feasibility of the designed system and method is verified by experiments, and short-term prediction can be achieved more accurately.