Fast Nonlinear MPC for an Overhead Travelling Crane

This paper presents a nonlinear model predictive control scheme for the two main axes of an overhead travelling crane, which guarantees both tracking of desired trajectories for the crane load and an active damping of crane load oscillations. The main idea of the used NMPC algorithm consists in a minimization of the tracking error at the end of the prediction horizon. That way the computation load can be kept relatively small. The varying length of the rope is considered by gain-scheduling techniques. The position of the crane load is measured by a CMOS camera using the spatial filtering principle. Desired trajectories for the crane load position in the three-dimensional workspace can be tracked independently with high accuracy. Experimental results from an implementation on a test rig show a high control performance.