Research On Active And Passive Driven Lower Limbs Powered Exoskeleton Robot

Exoskeletons are wearable robots which can effectively improve the wearer's mobility.They find extended applications mainly in military,industrial,and medical domains.As the subject of this research project,the light-duty exoskeleton has both active and passive mechanisms to aid its wearer in military combats and industrial handlings,achieving the goal of reducing or even eliminating the heaviness felt by the user when loaded with weights.This report discusses the mechanics,control and detection systems,control strategies,relevant experimentation and results of the exoskeleton with an emphasize on its applications.As for the mechanical structure,ergonomics and relevant human motion data were referenced to meet the anthropopathic design.In the premise of bearing the target load,the weight and dimensions of the exoskeleton were reduced as small as possible.The exoskeleton utilizes an active and passive combined drive mode,balances itself with a four-bar linkage mechanism with variable gear ratios active joints to reduce the weight and the force generated great,thus realizing the goal of eliminating the weight felt by the wearer.The hardware used in the control system is composed of a main control unit with DSP and ARM.The distributed control system was designed based on CAN.Aiming at the light-weight system design principle,Hall Effect angle sensing modules were implemented for posture detection and the establishment of the control model.As parts of the kinematics detection module,flexible pressure sensors and contact sensors were installed at the bottom of each foot to effectively sense the wearer's motion.In the drive unit,force-sensing modules with two measuring ranges were designed for moving intention identification.The application of the control algorithm is ensured with simple structures and high accuracy.As for the control strategies,a coordinate system was established for the exoskeleton and the kinematics were analyzed.Newton-Euler method was applied to model the dynamics in order to each locations and each joints.Linear 3D inverted pendulum model was utilized in support phase for gravity compensation,along with sensitivity amplification,to achieve a favorable pursuit movement and support.In swing phase,sensors with two measuring ranges were used to detect the interaction force between the wearer and the exoskeleton,to predict the wearer's moving intention and to provide feedforward prediction compensation such that the movement can be followed instantly.The exoskeleton was assembled and integrated to conduct relevant experiments.The joints were tested individually for their functionalities.Active joints were tested for passive joints were tested for their balance effect.Furthermore,squatting with loads,continuous walking,mobility of the exoskeleton,complex walking conditions were all tested to successfully verify the support of the exoskeleton.The results are as follows: own weight 15 kg,maximum load 70 kg,maximum walking speed 5km/h,which suffice the design requirements.