Nonlinear model predictive control of hydraulic forestry crane with automatic sway damping

Abstract

Forest machines are used in many tasks and come in various designs. They can be used for cutting down trees, collecting logs and in different forest cleaning operations. Currently, in the commercial machines the level of automation is still relatively low, and they require a professional operator for good work efficiency. Nonlinear model predictive control (NMPC) is an optimal control strategy based on a dynamic model of the system. NMPC algorithms require quite a lot of computational power, but are becoming a more viable option as the performance of computers has increased. We demonstrate how an NMPC can be used for controlling a hydraulic forestry crane that has a freely hanging tool or processing head attached. The goal of the control is to follow a predefined path while simultaneously damping the undesired oscillations of the tool. Three different reference paths with velocities of 0.5m/s to 1.0m/s are tested. The average tracking error in these tests is between 0.02m and 0.11m. Anti-sway control can reduce the amplitude of sideways oscillations between 2% and 64% and longitudinal oscillations between 59% and 76%. The impact of anti-sway control on the tracking accuracy or the velocity is negligible.