Model reference input shaped neuro-adaptive sliding mode control of gantry crane

Abstract

This article proposes a hybrid control strategy that integrates the merits of both zero-vibration double derivative input shaping control and neuro-adaptive sliding mode control. The approximation ability of the radial basis function neural network is utilized to estimate the unknown dynamics of the gantry crane system. The input shaping control is used to shape the trolley position, velocity, and acceleration references. These reference trajectories are generated as the state trajectories of a closed-loop system obtained by controlling a nominal system model with a sliding mode control constructed based on the nominal model. The shaped trajectories are sent to the proposed neuro-adaptive sliding mode control as references to be tracked. In comparison with the conventional sliding mode control, the proposed control has higher tracking accuracy and robustness to external disturbances, and effectively reduces the sway angle without suffering from the chattering phenomenon. Moreover, the proposed control has a very high insensitivity to parameter uncertainties. The adaptation mechanisms of the radial basis function neural networks are established based on the Lyapunov stability theory. The effectiveness of the proposed controller is verified through some numerical simulations.