A physics- and data-based method for sloshing active control

This study presents an experimentally validated physics- and data-based approach to optimize piston-type actuator amplitudes in shipboard pools for active control of sloshing-induced wave elevations under resonance conditions. The piston actuators, installed at both ends of a pool, are driven periodically to counteract waves generated by the pool's motion. Building upon our previously validated linear numerical method that accurately predicts non-resonant wave dynamics using actuators, we established an efficient framework to determine optimal actuator thresholds. Experimental investigations in two scaled pool configurations at the sloshing laboratory of the University of Duisburg-Essen confirmed the numerical method's effectiveness in identifying critical actuator amplitudes for resonance mitigation, despite inherent limitations in modeling nonlinear wave phenomena. To address nonlinear characteristics in resonant regimes, we implemented a data-driven Kolmogorov-Arnold Network trained on experimental wave profile data. This hybrid method demonstrated exceptional computational efficiency, enabling rapid optimization of wave mitigation systems using piston-type actuators. Experimental results validated both the reliability and the practical feasibility of the proposed hybrid method to optimize piston-type actuator motions for resonance control in confined waters. This work established a baseline method for active sloshing control system design, combining computational efficiency with experimental verification. The proposed approach bridged linear system analysis with nonlinear dynamic responses through synergistic physics-data integration, suggesting pathways for enhanced nonlinear coupling analyses in future research developments.