Nonlinear Control of Ship-Mounted Rotary Crane Based on Adaptive Dynamic Programming

As a crucial component in marine transportation, the precision of ship-mounted rotary cranes directly impacts the efficiency and safety of lifting operations. However, the inability to directly control the swing angle of the load through the drive mechanism renders ship-mounted rotary cranes inherently complex, featuring underactuated and coupled characteristics. In this study, we propose an optimal feedback controller leveraging adaptive dynamic programming (ADP), which integrates sliding mode control with optimal control strategies. Specifically, we devise a judicious cost function and solve the Hamilton-Jacobi-Bellman (HJB) equation using the adaptive dynamic programming approach. Subsequently, the optimal feedback controller is derived through iterative batch neural network training facilitated by the adaptive update algorithm. Moreover, we establish the stability of the proposed controller through rigorous Lyapunov techniques and LaSalle’s invariance principle. Experimental validation confirms the efficacy of our approach, highlighting its practical utility in real-world scenarios.