Weather route planning for ships based on improved grey wolf optimization

As the shipping industry transitions toward intelligence, weather route planning is essential for energy efficiency, operational performance, and safety. The Grey Wolf Optimizer (GWO) is widely used in path planning, but its application to weather routing suffers from inadequate balance between global and local search and premature convergence. To improve search accuracy and adaptability for ship routing optimization, this paper proposed an improved GWO that integrates golden sine search and dimension learning-based hunting strategy (GSDLHGWO). The golden sine mechanism enhanced population diversity and provided macroscopic directional guidance, while an adaptive neighborhood learning scheme performed microscopic dimension-wise corrections to balance exploration and exploitation. The improved algorithm was embedded into the Pareto multi-objective framework to obtain trade-off solutions among multiple objectives. The statistical results of single-objective experiments across multiple scenarios demonstrated that GSDLHGWO achieved substantial and statistically significant fuel savings across most scenarios while remaining competitive in voyage time minimization. In multi-objective tests, the algorithm yielded wider-coverage Pareto fronts with better diversity, producing multiple balanced and feasible routes. Overall, the proposed algorithm exhibits superior optimization capability, providing effective technical support for safe and efficient route planning in smart shipping.