| With the improvement of the requirements in lower limb rehabilitation training,the lower limb rehabilitation robot equipped with body weight support system has shown benign rehabilitation effect in clinical application.In recent years,with the improvement of rehabilitation needs as well as investment in related research,body weight support system not only maintains the original function of body weight support,but also gradually integrates the functions of Co M detection&control as well as posture control.However,most of current lower limb rehabilitation robots still have problems of long-time man-machine combination process,low degree of automation,complex operation,and insufficient detection means for Co M and posture control.Aiming at the above problems,this paper puts forward the following innovative work:(1)A lower limb rehabilitation robot(LLRR)with convenient human-machine combination,capacity of Co M motion in multi-dimensions and posture control is proposed.According to pattern of Co M and posture during walking,LLRR that compatible with activeand-graded body weight compensation,Co M control in two-dimensions,posture control and 3-Do F lower limb exoskeleton is proposed,which can automatically adapt to patients with different sizes.As the core,the body weight support system(BWSS)is supported by singlearm rotary support structure(SARSS).BWSS ultilizes the characteristics of quick passive response of the spring group and the active suspension to realize the reciprocating motion of Co M in vertical axis,and realizes the reciprocating motion of Co M in coronal axis as well as shoulder rotation by a 2-Do F mechanical arm.It relies on SARSS to reduce the space obstacle for patients to enter the rehabilitation platform,and realizes the rapid combination between patients and LLRR by the function of automatic adjustment as well as mechanical fast wearing interface.(2)An optimization scheme for BWSS that based on adaptive multi-objective optimization and NLPQL algorithm is proposed.A method based on the shape and inertia moment of section is adopted to simplify the 3d models of BWSS and SARSS.The results of FEA based on ANSYS show that BWSS and SARSS that constructed by industrial aluminum profiles have problems of excessive deformation and structural unsafety.Therefore,an optimized structure was proposed,and four key parameters were optimized by NLPQL algorithm based on adaptive multi-objective optimization.The maximum stress of the optimized structure was reduced by more than 72.6%,the maximum deformation was less than 1/150 of the cantilever,and the weight of the structure was minimized.(3)A contactless Co M perception approach based on Kinect is proposed in this paper.In order to meet the requirements of Co M detection in lower limb rehabilitation training,Kinect was used to measure the length of different limb segments of human body,then the centroid position of each segment was calculated according to the empirical parameters,and finally the position of Co M of whole body was calculated.A 9-bar human body model was established in MATLAB,whose simulation results show that and the simulation curve of Co M had the same motion characteristics as the Co M trajectory in natural gait,which verified the feasibility of this perception approach.The results of walking experiment show that the maximum measurement error of Co M is less than 4.03%,which indicates that the proposed approach has advantages of considerable ability of Co M measurement.The contactless measurement is simple to operate and has great cost advantage and clinical application potential.In conclusion,the LLRR and its BWSS proposed in this paper,combined with the perception function of Co M,can solve the problems in the existing lower limb rehabilitation robot,and has a broad prospect of clinical application. |
PDF Full Text Request