Design Of Parallel Rehabilitation Robot System Based On Human Motion Analysis

In recent years,the aging of Chinese population is becoming increasingly serious.The number of elderly people with lower limb motor dysfunction caused by cardiovascular and cerebrovascular diseases such as cerebral stroke is increasing year by year,and nearly 80% of them are in urgent need of rehabilitation diagnosis and treatment.In addition,the number of patients with walking disorders caused by traffic accidents and sports injuries is also very large.In order to ensure the basic walking function of patients to meet the needs of daily life,it is necessary to carry out scientific rehabilitation training for these people.At present,sports rehabilitation training mainly centers on rehabilitation physicians,but due to the small number of rehabilitation physicians in China and the large number of patients,it is impossible to guarantee the training time and intensity of each patient.In order to solve the problem of lack of training caused by the shortage of rehabilitation physicians,more and more researchers are devoted to research rehabilitation robots.Firstly,this paper analyzes the motion state of pelvis in healthy human gait,combined with pelvic kinematic data from the experiment and the hip width and hip thickness of the subjects,obtain the real-time position of the pelvis model,and analyzes the mechanical characteristics of plantar in the process of human walking,which provides the bionic basis of human motion for the mechanical mechanism design and safety evaluation of rehabilitation machine.The rehabilitation robot is a kind of humancomputer cooperation equipment to assist walking.In order to judge the walking intention of patients and enhance the performance of human-computer interaction,a gait phase identification system is designed.The wearable inertial sensor is used as the motion acquisition unit to collect the kinematics data in the process of human walking in real time and obtain the joint angle data through calculation.The gait cycle is divided into four stages based on the data characteristics of joint angle,and the Support Vector Machine model is selected to achieve the classification and identification of gait phase.Based on the bionic data of human motion,we design a parallel rehabilitation robots by controlling the independent telescopic movement of three parallel rods,driving the adjustment of the saddle position and posture,thereby driving the human pelvis to achieve pose synchronous changes,and through the differential between the front and rear driving wheel rolling,implements the tensioning movement of chassis,guarantees the demand of passing through narrow space in the process of indoor use.In order to evaluate the safety performance of the equipment,finite element software is used to analyze the stress condition of the equipment in the limit state,and d’ Alembert principle is used to analyze the dynamic stability of the structure.Finally,the relationship between the position and speed between the saddle and the drive rod is analyzed through kinematic modeling,and the control system of the device is further built based on MATLAB.The rationality of the parallel rehabilitation robot is verified by comparing the output position and posture of the saddle with the bionics data of human movement.