Research on Multi-Objective Optimization Method for Ship Tank Loading Based on Improved Non-dominated Genetic Algorithm

Ship tank loading and unloading is one of the most frequent and typical operational scenarios. Meanwhile, ship loading and unloading operations are influenced by various factors such as meteorological and hydrological conditions in the operational waters, operational safety redundancy, operational windows, and more, all of which place extremely high demands on the efficiency and reliability of the loading plan. To improve the efficiency of ship tank loading and enhance control over the safety of the loading process, this paper constructs a complex loading condition characterized by calibrated loading capacity and floating state. An optimization model for ship tank loading plans is proposed, aiming to optimize the strength of the loading state. Based on the characteristics of this model, multiple optimization objectives and boundary conditions are sorted out. The NSGA-III method adapted to the specific needs, is used to facilitate the iteration of effective populations, thereby enabling the rapid generation of loading plans. This approach changes the traditional process of generating loading plans through multiple interactive iterations of first loading and then verifying against loading targets. Instead, it first determines the loading targets and then uses algorithms to find the optimal loading plan, improving the purposefulness of the loading work and optimizing the search process for loading plans.