Research On Force/Position Control Strategy For Multi-DOF Of Parallel Rehabilitation Robot

Parallel robots, due to advantages in strong load capacity, high stiffness, highaccuracy, compact structure, good control performance and dynamic performance, arebeing widely used in analog movement, movement docking, bearing exercise, etc.Combined with the structural characteristics of the parallel robot, Parallel robot isapplied in the field of rehabilitation of lower limb. Effectiveness and pertinence oflower limb rehabilitation can be improved by using reasonable control strategy andrehabilitation methods.Parallel robot has a strong coupling property, in order to achieve a betterrehabilitation treatment, effective rehabilitation strategies are needed, besides, thedynamic model of parallel robot, motion control strategy, and forward kinematics alsoneed to be studied. To solve the above problems, this paper will study the followingfour parts:(1) For the problem of positive solution is not easy to solve, fuzzy neuralnetwork model is used to establish forward kinematics of parallel platform, and iscompared to Newton-Euler method by MATLAB.(2) For the problem of dynamic modeling is complex, system identificationmethods is used to build mathematical model of the servo motor, SolidWorks andMATLAB are combined to get a more accurate dynamic model of parallel platforms.(3) For movement control strategy, model predictive control method is studied,the mathematical model used in this method which is derived from identificationcontrol algorithm. Traditional PID algorithm is also compared to MPC.(4) For lower limb rehabilitation exercise strategic issues, a force sensor isintroduced to study the impedance control algorithm and hybrid force/position control.Several rehabilitation exercise strategies are proposed and verified the feasibility ofrehabilitation exercise strategy on the actual test platform.According to the above research, we can build the mathematical model moreefficient. Combine with the forward kinematics, we overcome the problem that theclosed-loop control of the terminal during the debugging. Integrate with force sensor,we can design a better controller to improve the rehabilitation effects with themethods that already been studied.