Adaptive finite-time extended state observer-based model predictive control with Flatness motivated trajectory planning for 5-DOF tower cranes

Abstract

This paper introduces a new method to control a 5-DOF tower crane (3DTC). By considering the 3DTC as a flat system, a time-optimal trajectory is proposed for the payload. System states and control signal references can be calculated based on the flatness theory. In addition, the 3DTC works in an environment containing many factors impacting control performance and the system states are hard to measure. An adaptive finite-time extended state observer (AFT-ESO) is introduced to solve these problems. With AFT-ESO, system states and lumped disturbances can be estimated accurately, facilitating the prediction for Lyapunov-based model predictive control (LMPC) when an accurate model is required. The LMPC takes advance of the second-order sliding mode control stability conditions as a strict constraint to guarantee the global stabilization of the closed-loop system. Finally, simulations based on the quasi-physical model are proposed to show the effectiveness and robustness of the proposed strategy.