Movement Pattern Recognition And Fatigue Analysis Based On EMG Signals From Lower Extremity Surface Of Human Body

Due to diseases,natural disasters,sudden accidents and increasingly aging population,the symptoms of amputation,hemiplegia,hemiplegia and neuromuscular atrophy have brought huge economic burden and psychological damage to the patients and their families and seriously affected their normal life.If these patients for targeted rehabilitation plan,let them a certain amount of rehabilitation training,improve their athletic ability and physical quality,making their body movement mechanism through rehabilitation training to normal levels,relieve the suffering of the patients,this will be the final home to return to the human body the multi-channel semg and foothold.In this paper,surface emg signals of human lower limb movements are systematically studied.Firstly,surface emg signals were collected for 7 motion categories of lower limbs,4 flexion and extension motion amplitudes and frequencies of knee joints in 2 postures,3 joint motion modes in sitting position and 2 knee flexion and extension fatigue in 2 postures.Then,starting from the time-domain analysis,frequency-domain analysis and nonlinear analysis,two kinds of eigenvalues were extracted by each analysis method,and six kinds of eigenvalues were extracted by the three analysis methods.Then,using BP neural network method of human lower limb seven categories for pattern recognition,with ELM limit adaline neural network method for two kinds of attitude under knee four flex movement amplitude frequency of pattern recognition,using BP neural network method for three kinds of joint movement posture condition modes,pattern recognition,get higher recognition rate,the real-time pattern recognition control experiment is completed.Finally,according to the extracted characteristic parameters,the fatigue degree of human movement was analyzed to find out the internal relation between the fatigue degree of human movement and the characteristic parameters of emg signal,so as to ensure the safety and reliability of human rehabilitation training.In this paper,the real-time active control of virtual reality rehabilitation training scene and electromechanical movement system is realized by using the results of real-time pattern recognition of human surface emg signals,and the connection between theoretical research results of human surface emg signals and practical application is completed,so as to truly achieve more intelligent and humanized rehabilitation training.By extracting the patient's body movements movement of human biological information,multi-channel semg as virtual reality rehabilitation training scene and electromechanical movement system real-time control of the source signal,control of virtual reality rehabilitation training scene and electromechanical movement system movement patterns,movement on the surface of the sport changed,patients with rich training way,increase the training in patients with gout,improve the efficiency of patient training,has the very important recovery value and practical significance.