The Design And Experimental Study Of The Assisted Handling Exoskeleton Based On SEMG

With the development of science and technology in modern society,more and more intelligent equipment has been used in the handling field,but manual material handling still exists widely in various walks of life.As wearable human-machine integrated equipment,exoskeleton enhances human body functions with the help of machines and reduces the body’s own energy consumption.The exoskeleton has unique application prospects in the fields of logistics,medical treatment and emergency rescue.This paper focuses on the research of the assisted handling exoskeleton,and analyses the manual handling process by human biology.Then the study focuses on the action recognition method based on the surface electromyographic(sEMG)and the system design of the exoskeleton.Finally,the overall performance of the prototype is evaluated through handling experiments.This paper carries out a biological analysis of the manual handling process.Based on the analysis of the motion characteristics of human skeletal joints and the types of musculoskeletal injuries,a human musculoskeletal model is established based on Opensim.Then the kinematics and biomechanical models of the handling process are studied.The mechanical relationship among lumbar spine tension,handling load and bending posture is constructed.Based on the generation mechanism and characteristics of sEMG,a signal acquisition and preprocessing method for sEMG is designed.The analysis of sEMG includes feature extraction and feature selection methods,and the classification principles of RBF neural network and SVM algorithm are discussed.The recognition effects of the two algorithms are compared with the data collected in the handling action recognition experiment.The results show that the SVM algorithm based on the RBF kernel function has the highest classification accuracy,and the overall classification recognition rate reaches 95.3%.According to the design principle of the assisted handling exoskeleton,the overall design of the exoskeleton system is carried out.The overall size and degree of freedom of the exoskeleton are determined.The design of the mechanical structure includes shoulder,back and lower limbs,and the flexible binding structure is adopted to wear and fix the exoskeleton.The control system uses sEMG as input signal and uses Arduino as control core.The system is driven by a motor combined with an electromagnetic push rod.ANSYS finite element simulation is performed on the back support with the largest load,and then the structure is designed and optimized based on the simulation results.Based on the design scheme of the assisted handling exoskeleton,a prototype of the exoskeleton is built and tested for reliability.The experimental scheme of exoskeleton performance evaluation is designed,and a comprehensive evaluation system combining physiological index testing and subjective evaluation is established.Compared with the test without exoskeleton,the sEMG amplitude of the wearer’s back and arm muscles decrease significantly,the heart rate and oxygen consumption indicators are reduced,and the subjective fatigue level is also reduced.The experimental results verify the power effectiveness of the exoskeleton prototype.The energy consumption and perceived fatigue can be reduced significantly by wearing the exoskeleton to carry out handling tasks.