STUDY ON DYNAMIC MODELING AND NEURAL NETWORK CONTROL FOR FREE-FLOATING SPACE FLEXIBLE-JOINT ROBOT TO TRACK DESIRED TRAJECTORY IN JOINT SPACE

摘要

In this paper, the dynamic and control problems of free-floating space manipulator system with general flexiblejoints are discussed. With the relationship of linear and angular momentum conservation, the dynamic equations ofthe system are derived by Lagrangian formulation. To overcome the shortage of singular perturbation approach thatcan only be utilized in the control of manipulators with weak joint flexibility, a joint flexibility compensator that caneffectively improve the equivalent joint stiffness is introduced. By using the singular perturbation theory, the systemdynamics model after exibility compensation can be easily decomposed into two subsystems: the fast subsystem andthe slow subsystem. Then, a neural network control scheme is proposed for the slow subsystem to track the desiredtrajectory in joint space, while the stability of the fast subsystem is guaranteed well by a torque differential feedbackcontroller. A simulation example is carried out to illustrate the validity of the proposed control scheme.