The Modeling Of Driving System For Lower Limb Rehabilitation Training Device

Rehabilitation training device is a new kind of medical robot appearing in recent years. Its security requirement is greater than traditional robot, so the more precisely control is needed by it. However the modeling of the driving system is the premise of mechanical control, so the Modeling of the System is the focus of the paper.Support vector machine is the latest model in the statistical learning model. It inherits the advantages of machine learning model and has the capability of fitting and generalization. It also has the unique advantage over the traditional model, just like mechanism model, half mechanism model and neural network.The major works of this paper are listed as follows:Firstly, through the simulation for sampling, coding and processing of drive system.Secondly, the support vector machine identification device for the modeling of the system is designed. And it also proved by experiments. The results of the experimental show that the model can effectively fitting system.Finally, three kinds of different algorithms for the identification device parameter optimization are designed. The experiment proves that each of the three models has their own advantages:1) Small sample particle swarm optimization model has a small amount of computing and good real-time performance;2) High accuracy optimization model has strong capability of fitting;3) Compromise optimization model fitting and generalizing ability are both strong.