Neural network robust controllers applied to free-floating space manipulators in task space

摘要

This paper deals with the problem of robust trajectory tracking control for free-floating manipulator systems subject to plant uncertainties and external disturbances. The system model is described through the Dynamically Equivalent Manipulator approach and the tracking problem is formulated directly in task space. Two adaptive techniques are developed considering nonlinear ℋ∞ controllers and artificial neural networks. Experimental results are obtained from an underactuated three-link fixed-base planar manipulator dynamically equivalent to a two-link free-floating planar space manipulator.