Research On Intelligent Control For Rigid And Flexible Space-based Space Robot System

With the development of space exploration, space robot will play more and moreimportant functions. Especially, the flexible space robot system becomes the majorsubject over the past two decades because of their advantages over rigid space robot.In order to ensure the operation of high accuracy, the vibration control of flexiblespace robot system must be considered during the controller design. In the presence ofunknown or uncertain system parameters unknown, the dynamics modeling, trajectorytracking controlling, and vibration suppression of rigid and flexible space-based spacerobot system are investigated.Firstly, the dynamic equations of rigid and flexible space robot system arederived by Lagrange equation and assumed modes method. It is found by analysis thatthe dynamic equations can be linearly dependent on a group of system inertialparameters.Secondly, the adaptive control, neural network fuzzy control, and robust-adaptiveneural network control are designed for dual-arm rigid space robot system toguarantee the trajectory tracking of base’s attitude, joint angles, and end-point. Inwhich, the adaptive control can overcome the effect of part unknown parameters. Theneural network fuzzy control doesn’t need the system parameters at all during thetrajectory tracking, while the robust-adaptive neural network control makes use of theknown system parameters better. In particular, it doesn’t require measuring theposition, velocity nor acceleration of the base because of an effective exploitation ofthe particular property of the system dynamics.Then, the control schemes of flexible space robot system without vibrationsuppression, using composite control, under singular perturbation theory, and usinghybrid control are proposed respectively. When the flexible vibration is not controlactively, the tracking error is very small by computed torque method and neuralnetwork fuzzy control, but there are large vibrations during the control. Usingcomposite control, a simple direct parameter adaptive rule is presented to suppress thevibration more quickly. Since this composite controller regard the flexible vibrationstabilizer as a perturbation of the whole system, the rigid motion controller obviously loses a certain degree of robustness. By singular perturbation theory, the singularperturbation model of the flexible space robot system is obtained. The flexiblevibration stabilizer and the rigid trajectory tracking controller can be designed intwo-time scale, it solve the superposition of the two controllers in theory. In hybridcontrol, the virtual control force concept is employed to generate a new hybrid desiredtrajectory reflecting the flexible vibration modes as well as the rigid desired trajectory.As the proposed sliding-mode control and sliding-mode neural network control trackthe hybrid desired trajectory, the flexible mode can be stabilized. The problem ofcontroller superposition is not appearing in the hybrid control.Lastly, the numerical simulations are carried out for all system controllers.Simulations results confirm that all proposed trajectory tracking control schemescan dominates the trajectory tracking for space robot system, and all vibrationstabilizers can suppress the flexible vibration actively and quickly in real time.