The Design And Control Of Lower Limb Exoskeleton Power Assist Robot

In recent years, because of the aging of society was serious and the pursuit ofhigher living standards, wearable power robot of lower limbs called more and moreattention in all areas of society. Power robot of lower limbs not only can be used inthe field of the military and medicine, but also can be applied to social relief andrescue work. This study is mainly aimed at solving the inconvenience and fatigue ofthe elderly walking, as well as solving the problem of the fatigue of long-distancewalking for ordinary people. The main objective of this paper is to design a set oflower limb power device, which can provide power for human to walk real-timethrough recognizing human movement intentions.Main research contents of this study including human gait analysis and experimentof lower limb of human, optimization of design and functional analysis of lower limbpower robot mechanical structure, identification and analysis of human movementintentions, dynamics analysis and modeling of robot, study of control strategy andsimulation of this device.1) In this paper, we analyzed the structure of human lower limbs and motion rangeof every degree of freedom, used CATIA to design mechanical structure of powerrobot of lower limbs and personified the structural design. Based on the kinematicsand dynamics analysis of human, we selected active and passive joint, and selectedactuate mode on the active joint.2) Use ADAMS to establish virtual model of the power assist robot, modifyparameters of the prototype, set constraints and revolute, and execute the kinematicsand dynamics simulations. The simulation result texted and verified the feasibility ofthe robot, optimized the model of the robot.3) Combined the previous model established by ADAMS with the control modelestablished in MATLAB/Simulink in order to realize the co-simulation of the system.Based on the characteristics of human gait data, we used PID control algorithm tocontrol lower limb power system, choosing human gait database data as the input of the system, achieving gait synchronization of the robot and the wearer finally.4) We analyzed the muscle groups of human, and chose the muscles related topeople’s walking as the source of sEMG.Used the Wavelet de-noising to process thesEMG, and preprocess the sEMG, after that analyzed the time and frequency domaincharacteristics, and extracted the characteristic value, and then used the method of theshort-time Fourier transform to get the pattern of the characteristic value. And then thestarting point of the movement was obtained. In the end, human movement intentionswill be chosen as the trigger signal of lower limb power device control system,achieving the design goals.The simulation results demonstrated that the control system had a good responsecharacteristic and a good ability of trajectory tracking，and its robustness was steady.What is more，it can provide an important theoretical basis for study of assist robot.