Fully Actuated System Approach-Based Neuroadaptive Control for Underactuated 4-DOF Tower Crane Systems With Unavailable States and Input Dead Zones

Inspired by the fully actuated system approaches, this article develops a novel neuroadaptive control method for 4-degrees of freedom tower crane systems under parametric uncertainties, unavailable state variables and dead zones. First, auxiliary variables that incorporate both actuated and unactuated states, along with their derivatives and proportional-differential terms, are constructed to reconfigure the nonlinear underactuated system into a linear fully actuated system, eliminating the need for any linearization operations. Subsequently, the radial basis neural networks are employed to deal with parametric uncertainties and dead zones. In addition, a state observer is established to tackle measurable states, avoiding the need for discontinuous robust terms and enhancing theprecision of compensation and positioning. Following this, the asymptotic convergence of the elaborately constructed auxiliary variables guarantees that both actuated and unactuated states precisely arrive at their desired values. As far as we know, this article offers the first fully actuated system-based continuous neuro-adaptive control solution for underactuated tower crane systems. A series of experimental results substantiate the effectiveness and merit of the presented control approach.