Research On Key Technologies Of Lower Limbs Rehabilitation Robot

A lower limbs rehabilitation robot is a kind of robot used for assist hemiple-gic or paralysis patients with walking difficulty caused by neuro diseases such as wind-stoke or brain injured to paractise walking after urgent stage.With the tradi-tional training methods, the therapists are very tired and it is difficult for patients to assure stabilization and persistence of the training. Robot technology has great advantage to realize the quantitative and repetitive motion input. Not only supply assistance motion from the beginning stage of the rehabiliton, which is more benefit to the occurrence of the function recombination or compensation in central neuro system, rehabilition robot can also supply multiple training modes, various scene environments and feedback the the training effectiveness in real time to sat-isfy the requirement of walking practice. It can be used in convalescent home and household.The robot has great practical value and wide maket prospect. This dis-sertation was sponsored by the Nation Natural Science Fund "Research on several key technologies about lower limbs rehablitive robot". The main research work covers the design of general mechanism and control scheme, analysis of the kinematics and dynamics, motion planning and control stragy of a noval limbs rehablition robot. The mechanism scheme, theoretical derivation, simulation model's building and analysising were demonstrated in detail.This paper reviewed the reseach status of the lower limbs rehabition and ra-lated fields domestically and abroad. Based on the analysis of gait motion of the human, the general mechanism scheme and control scheme with three training modes of a novel lower limbs rehabition robot were bringed forward. The robot is consisted of a 4-DoF gait mechanism and a 5-DoF pelvis mechanism and they can be control harmoniously to generate three dimension gait motion of human to meet demand of rehabilitatiom training for patient.The forward and inverse kinematics of the gait control mechanism, pelvis control mechanism and 7-bar human body model with 12-DoF were analized, and then their mechanism models were build. By simulation, the corrections of kine-matics analysis and mechanism model were verified and build the foundation for the following dynamics analyses on these mechanisms. Considerring the characteristic of pelvis mechanism with two parelle branch chain, an analysis method to find the mechanism workspace was presented, which is also fit in the similar mechanisms to solve their workspace. Based on this method, an actual workspace was determind that can meet the demand of pelvis motion in the gait training。According to motion characteristics of the foot and pelvis when human walking in treadmill, the motion planning method of gait mechanism was pre-sented, and the paramaterized pelvis model and the optimizing method of the cor-responding parameter were presented. The human sticks maps in three vertical plane were draw and the hip, knee, ankle joint angles yieled were compared with the statistical gait joint angular data, and the rationality of above motion planning was checked.In order to analyzing the mechanism dynamics and control system, the Sim-Mechanis model of gait mechanism and pelvis mechanism were build. By kine-matics simulation, the correction was verified of the SimMechanis model and forward and inverse kinematics analysis of each mechanism. Their dynamic SimMechanics models with load were presented. The inverse dynamic simulaiton analyses of each mechanism were made and the desired motor and reducer pa-rameters were determined.The control system model of each mechanism were build in which the plant is the above dynamic SimMechanics models with load. According to load and re-quirements, the strategys of each control system were presented, and each con-troller parameters were adjusted by the LTI control analysis tool and singnal con-strait block.To verify the validation of the the motion planning and control systems, the passive human 7-bar solid SimeMechanics dynamics as the the real load, were integrated with the gait and pelvis mechanism control. The simulation results show the motion planning and control system were rational basically, meanwhile supplied useful reference data for developing the prototype and build foundation for the following research on control strategy of human and robot interaction.