Research On Mulit-modal Information Feature Fusion Method For Exoskeleton Robot

With the aging trend of the population,more and more elderly people are paralyzed due to diseases.Most paraplegic patients need to use exoskeleton robots to complete rehabilitation,but when the subjects use the lower limb to recover exoskeleton walking,they need to control the action of exoskeleton through physical keys.In this way,the hands used to support crutches need to operate physical buttons at the same time.This control strategy relying on manual operation not only brings low human-computer efficiency,but also increases the risk of misoperation.Therefore,in order to make the exoskeleton control system can bring higher man-machine efficiency and improve the fault tolerance of the control strategy,this paper proposes that the exoskeleton control system should be built around the human EMG signal and joint angle signal.The system should be able to control the exoskeleton according to the movement intention of the wearer,not based on the operation of the subject.In order to solve this problem,an action recognition system based on EMG and angle fusion is constructed to control the action of exoskeleton robot.The main content of this paper is divided into control system platform building,model building,experimental verification and analysis.The proposed fusion model(EA-CNN)achieves more than 90% of the accuracy of motion recognition,and the model is compared with a variety of machine learning recognition models to verify the accuracy of the proposed model.At the same time,the paper also completed the online verification experiment with the existing exoskeleton robot.In the experiment,the experimental conditions,experimental paradigm and related comparative experiments are explained in detail.Experiments show that the exoskeleton control system and mulit-modal fusion method can effectively help the disabled(experimental subjects)to walk without external physical control.Finally,combining with the existing conditions,the method of how to ensure the exoskeleton robot can achieve certain stability in the control system is given.