Research On Motion Pattern Recognition And Control Of Lower Limb Aid Device Based On Data Driven

Exoskeleton is an electromechanical device for enhancing,helping,and restoring the body’s athletic ability.The lower extremity exoskeleton assisting device can provide assistance for the elderly and patients with lower limb disabilities in their daily walking.With the acceleration of the aging process of the population,the study of the knee exoskeleton of the lower extremity is of great significance.In this dissertation,the design of terrain recognition system,the analysis of lower limb joint angle information and the selection of control methods are studied.The main contents of the dissertation are as follows:Firstly,physical signals such as speed and acceleration of the human body are generated along with the motion process.Collecting motion data to identify human intention has a certain hysteresis,which easily leads to the inconsistency between the body and the exoskeleton movement.In order to solve this problem,a recognition system based on two-dimensional laser scanning rangefinder is designed in the dissertation.The terrain environment information is collected by the sensor installed in the waist position of the body,and the learning vector quantization method is used to identify five different road conditions.After pre-identifying the human body’s intention,the control method can be developed before the exoskeleton drives the actuator,and then the gait is adjusted to achieve the best coordination effect between the human and the exoskeleton movement to ensure the safety of the exercise.Secondly,the flat ground,the up/down stairs,and the up/down slopes are five common terrains in life.The VICON MX three-dimensional gait information acquisition platform is used to collect the motion information of the joints of the body under different terrains.The acquired angle information is discrete information.In order to provide a reasonable reference trajectory for the controller,the Fourier series fitting method is used to fit the discrete joint angle information.Finally,the modeling of human lower limbs and exoskeleton is complicated and inaccurate.A data-driven model-free adaptive control method is proposed to control the lower limb joints.A dynamic linearization model is established based on the joint angle of the input and output data.The simulation prior torque is introduced to improve the accuracy of the controller by using the data trend of the torque when the human body is walking normally.Using Solid Works to design the human body model,the co-simulation ADAMS-MATLAB platform was built.The experiment was carried out by selecting the flat ground conditions.The simulation results show that the human gait wearing the exoskeleton is normal.Data from these assessments indicate that the proposed strategy enables the knee exoskeleton to track the trajectory of angle well and has a good performance on walking assistance.