Research On The Design And Control Of Flexible Ankle Exoskeleton

At present,the lower extremity exoskeleton generally adopts the integrated design of hip,knee and ankle.the heavy mass,large volume,and the mechanical structure of traditional rigid exoskeleton become a burden when there is no external energy input,leading that the traditional rigid exoskeleton has low efficiency,weak endurance,poor wearability and can not provide customized assistance for the lower limb joints.Therefore,it is meaningful to design a new type of modular lightweight flexible single joint lower extremity exoskeleton that can provide assistance based on the physiological characteristics of the lower extremity joints.Because the ankle joint provides the main energy for human propulsion during human walking,the design of the ankle joint assisted exoskeleton plays a crucial role in the study of modularized lightweight flexible exoskeleton.In this paper,the mechanical structure,high-level control strategy and low-level control algorithm of flexible and lightweight ankle exoskeleton were designed and studied,and the effectiveness of the work was verified by simulation and experiment.In terms of mechanical structure,two sets of flexible exoskeleton design schemes for ankle joints were proposed based on human biology.Considering the way of applying force,the length of the arm and the space occupancy,the second longitudinal force-exposed exoskeleton,which is driven by a rope spring,was finally determined.,while the output power of the exoskeleton motor was reduced 52.8% by optimizing the spring stiffness.In the aspect of high-level control,in order to judge the timing of the assistance,the characteristic value of calf inclination during human walking was identified through the research and observation of experimental data,and the time domain control strategy based on IMU was proposed.It can judge the movement state of the human body in advance by 0.034 S,and the effect was good,which can replace the function of the sole pressure sensor used in traditional exoskeleton,and reduce the resulting damage of the human motion intention misjudgment caused by the long-term use of the foot pressure sensor.For the low-level control,for the purpose of performing the high-level control output stably,accurately and quickly,based on the flexible exoskeleton characteristics,we got the optimal underlying control algorithm through design and simulation based on speed and torque compensation control,Kalman Filter based Speed and Torque Compensation Control,Kalman Filter and RBF n eural Network based Speed and Torque Compensation Control.In terms of experiment,this paper completed the construction of an ankle joint flexible exoskeleton experimental system,including hardware system and software system.In order to ensure system security,software limit was designed.The correctness of the spring stiffness design proposed in this paper was verified by the spring stiffness optimization experiment.Through the high-level control experiment,the accuracy of IMU-based time domain upper layer control gait division was verified.Through the control experiment of the low-level control,it was verified that the control effect of Kalman Filter and RBF Neural Network based Speed and Torque Compensation Control was optimized,and the error can be minimized to 0.8 Nm.Finally,through the overall performance evaluation of the flexible exoskeleton of the ankle joint,it was verified that the lightweight flexible drive exoskeleton can also provide a large assist torque.The quality of the ankle joint flexible exoskeleton implementation device was significantly lower than that of the rigid exoskeleton.Compared to the existing ankle joint flexible exoskeleton design,the end effector mass was 22% lower than the CMU Alpha exoskeleton and 38.7% lower than the V2 exoskeleton with the same assist torque.