Design And Research Of Exoskeleton Robot Based On EMG Signal To Correct Foot Drop

Foot drop causes the patients are unable to walk normally.Their exercise ability can be improved through rehabilitation treatment.Rehabilitation robot assisted foot drop patient training is one of the rehabilitation methods.However,most of the robots used for foot drop rehabilitation training can not assist the human body to complete the coordinated movement of man-machine synchronization according to the wishes of patients.In order to solve the problem of motion fusion and synchronization between man and machine,an exoskeleton robot with human will as motion navigation information is designed..Because surface EMG(s EMG)signals can predict human movement intention,in order to ensure the accuracy of prediction,a method of navigating exoskeleton robot by using the rules of s EMG signals between different muscle groups is proposed.According to the muscle parts driving the skeleton of human lower limbs,the acquisition area of s EMG signal is selected,and finally it is determined that the acquisition area is located at the thighs and shanks in human body.The s EMG signal is collected in a variety of noise environments,and the change of s EMG signal acquisition accuracy is analyzed,.On the premise of ensuring the acquisition accuracy,the s EMG signals of healthy subjects and patients with foot drop are collected.By exploring the relationship of s EMG signals between muscle tissues,the rules of s EMG signals between thighs and shanks during human walking are found.Based on the design principle of rehabilitation robot,a flexible exoskeleton robot is designed.The positive solution model of the exoskeleton robot is established,and the inverse solution equation is derived,which verifies the rationality of the geometric design of the exoskeleton robot.In order to ensure the safety of subjects,the dynamics of exoskeleton robot is analyzed.The stress analysis and parameter design of the important stress parts of the exoskeleton robot are carried out,the feasibility of exoskeleton robot driving human walking is verified.The s EMG signal is preprocessed and the time-domain eigenvalue is extracted.According to the explored s EMG signal law and eigenvalue,the BP neural network training model is established.The training results are analyzed,and the accuracy of BP neural network in motion pattern classification is verified,and the feasibility of navigating exoskeleton robot by s EMG signal law between human thighs and shanks is proved.In order to realize man-machine communication and master the fatigue degree of patients,a man-machine interface is designed,then active rehabilitation training strategies and passive rehabilitation training strategies are formulated.The prototype of exoskeleton robot is built.Continuous no-load experiments were carried out on the prototype to verify its safety.Human gait experiments and rehabilitation experiments in sagittal plane were carried out to verify that the explored s EMG signal law can guide the exoskeleton robot to drive the human body to realize normal walking.Moreover,a virtual scene is established to help patients with foot drop carry out various forms of rehabilitation training.