A Backstepping Control Method For Lower Extremity Exoskeleton Based On Extended State Observer

Since the beginning of the 21 st century,with the rapid development of science and technology,robotics has also developed rapidly as a technology that integrates multiple disciplines.Among them,the lower limb exoskeleton robot is connected to the human body through wearing and can control all joint movements.So it can be widely used in the military field and the rehabilitation treatment field.The lower limb exoskeleton robot needs to use the information acquisition module to collect the human-machine interaction information of the operator in real time.The effect of the control strategy directly determines the performance and safety of the exoskeleton robot.Therefore,it is of great significance to carry out research on the control method of the lower limb exoskeleton robot.This thesis designs and develops a lower extremity exoskeleton experimental system,including the software and hardware design of the robot platform and the design and implementation of the lower extremity exoskeleton control algorithm.The usability of the control algorithm was successfully verified.First,the second-generation double-leg lower-limb exoskeleton robot experimental platform,named UEXO-II,was designed and built.The mechanical structure conforming to the human body is designed for the UEXO-II experimental platform,and various measures such as plastic hard protective layer,sponge cushion and joint hard limit are adopted to ensure the safety of the wearer;the UEXO-II experimental platform is equipped with a set of sensing and driving systems are used to realize the precise collection of joint driving and human-computer interaction information;in order to realize the collection and processing of the human-computer interaction information and the real-time control of the joints,a set of host computer software platform is designed on the high-performance host.It can be used to control and test the experimental platform in real time.Then,in order to achieve the goal of "host-assist" to help the human body to perform rehabilitation training in accordance with the motion trajectory,a lower-limb exoskeleton controller which can assist human motion on the premise of ensuring safety and comfort is designed.Aiming at the influence of inaccurate parameter identification and unknown external interference in the actual control process,a controller with certain antiinterference ability is designed.At the beginning of this chapter,it introduced the relevant theoretical knowledge.On this basis,a backstepping controller based on Nonlinear Extended State Observer（NESO）is designed,the stability of the controller is proved,and simulation experiments are carried out.Finally,the designed backstepping controller based on NESO is applied to the UEXO-II experimental platform to verify the control algorithm.Before the experiment,the system was debugged in various aspects to ensure the effectiveness of the system.After that,the human gait trajectory was designed as the input of the system,and the appropriate parameters were selected for the controller and the NESO nonlinear extended state observer according to the parameters of the actual platform.At the end of this chapter,the experimental results are analyzed in detail,which proves that the UEXO-II experimental platform has high reliability,and also proves that the NESO-based backstepping controller can ensure the stability of the closed-loop system and has certain anti-interference characteristics.