Lower limb rehabilitation robot is rehabilitation training intelligent medical equipment which was used auxiliary lower limb functional dyskinesia patients. In view of the lower limb exoskeleton rehabilitation robot is studied. The main content includes the following several aspects:(1)According to research by this group of lower limb exoskeleton robot functional requirements,And the characteristics of the patients with lower limb rehabilitation training,appropriate human gait planning method is proposed. For specific patients respectively use off-line gait planning parts and fit for the actual online adaptive adjustment of gait training parts. Offline part is to provide Rehabilitation doctor for patients a more accurate and objective scope of gait( Step length, the gait cycle) to guide the patients in the most comfortable and natural condition to walk normally Online planning part is Rehabilitation doctor according to reference data for patients in the process of rehabilitation training on appropriate adjustments. Using this method is suitable for individual differences of patients and can be got gait planning which make patients get better effect of rehabilitation training.(2)Offline part is this process which according to patients own physical parameters and gait training program plan patients training. By using RBF neural network for training the sample data obtained,six Body parameters as training input(age, height, weight, sex, thigh length, leg length), the output is the need to step length and the gait cycle.(3)Online planning is when patients according to the gait planning training and under actual gait produced skewed gait, adaptability of gait planning immediately, make the new gait, which can adapt to the patient’s actual gait and conform to the normal gait characteristics, meanwhile avoid training injury for patients, have prevent patients with pathological gait, influence the training effect. Online planning part adopts an adaptive CPG generate real time gait trajectory of lower limb exoskeleton method. By according to the step length and the gait cycle to solve the joint Angle trajectory, and its fitting is a limited order Fourier series. Then the fuzzy relation was established between Fourier coefficient and step lengthã€the gait cycle. After, the establishment of CPG network and that is associated with Fourier coefficient, form a real-time online gait planning.(4)Based on the research and development of lower limb exoskeleton rehabilitation training robot model. The joint simulation was established between Solidworks and matlab/simulink. It can realize off-line prediction and online planning, the feasibility of the method is validated. |