Non-linear Robust Adaptive Control Of 6-DOF Parallel Robot

In the dissertation, first, we simply introduce the physical structure of parallel robot and the development of robot control theory; second, we make research onjsyslem simulation tool, after long time consideration, finally we found the best way to simulation is to comply Simulink and S-function, it can avoid the complicated process to transfer nonlinear robot dynamic model to state-space model; then, in the third chapter, we mainly make research on the robot theory, on how to design robot controllers about general series robots; finally, we focus on the six degree parallel robot model made by Yanshan University, on the basis of former postgraduate's work, we present the uncertain parallel robot model, then prove three important robot qualities, after that, we imply robust-adaptive control decentralized strategies which depend on neither the nominal model nor the regressor into parallel robot system as the controller, which includes a simple PD linear controller and a nonlinear feedback to compensate the uncertain dynamics. The proposed controller can get rid of the effects of uncertainties and the only information required in setting up the strategies is the position and velocity state of each joint. In addition, the decentralized controller not only considers fully the dynamic interaction of each joint and effects of unmodelled dynamics but also overcomes the unsmooth controller output and uncontinuous structure of existing decentralized controller and can guarantee the global asymptotic stability. And also, as to the proposed robust-adaptive decentralized controller, we did a large amount of tracking simulations. The results indicated that the strategies presented in the dissertation was theoretically correct, it can meet the designing demand of the robot systems.