Editorial
Integrating model predictive control with deep learning for sway reduction in ship-to-shore crane operations
@ARTICLE{10.4108/tsoe.10838, author={Xuan-Kien Dang and Viet-Dung Do and Ngoc-Truc Nguyen and Soi Ly}, title={Integrating model predictive control with deep learning for sway reduction in ship-to-shore crane operations}, journal={EAI Endorsed Transactions on Transportation Systems and Ocean Engineering}, volume={1}, number={1}, publisher={EAI}, journal_a={TSOE}, year={2025}, month={12}, keywords={deep learning, model predictive control, prediction model, ship-to-shore crane, sway angle}, doi={10.4108/tsoe.10838} }- Xuan-Kien Dang
Viet-Dung Do
Ngoc-Truc Nguyen
Soi Ly
Year: 2025
Integrating model predictive control with deep learning for sway reduction in ship-to-shore crane operations
TSOE
EAI
DOI: 10.4108/tsoe.10838
Abstract
The sway of the container winch drive system results in significant nonlinearity in Ship-to-Shore (STS) crane operations. As a result, achieving an accurate winch path becomes challenging, raising safety concerns for the operator and increasing the risk of accidents to both goods and equipment. This paper presents a Deep learning-based Model Predictive Control (DMPC) designed to improve the precision of the winch control signal during STS operations, ultimately reducing the load sway amplitude. First, a Long Short-Term Memory (LSTM) is employed to compute a state prediction model that forecasts the load sway angle and winch displacement amplitude. The predicted state serves as input to the DMPC controller, which determines the winch control value through an optimization function. Finally, the proposed solution is tested through two scenarios, yielding promising results that demonstrate its effectiveness.
