Research on Multi-Objective Optimisation-Based Task Package Division Methods for Shipbuilding

Shipbuilding requires converting multi-disciplinary product units—such as hull structures, outfitting equipment, and coating processes—into executable task packages to support sectional construction and final assembly on the slipway. Three critical issues arise during practical division: firstly, weak inter-process connectivity within task packages, where splitting welding and assembly tasks within the same compartment disrupts workflow continuity and increases redundant handling costs; Secondly, high coupling between task packages arises when construction packages are empirically divided, placing sequentially dependent tasks under different crews and triggering frequent cross-departmental coordination. Thirdly, resource allocation imbalances occur when critical resources like machining equipment are concentrated or scarce, constraining the construction cycle. Therefore, this paper constructs a multi-objective optimisation mixed-integer programming model. First, leveraging information entropy theory, we quantify the homogeneity of shipbuilding processes within task packages—such as hull welding and piping pre-installation—to ensure process continuity and specialised concentration. Second, a task dependency matrix quantitatively assesses and reduces inter-package coupling, minimising cross-package coordination costs. Finally, an equitable resource allocation metric is introduced, using variance in critical resource utilisation to achieve balanced distribution. Simultaneously, linearisation techniques address non-linear constraints, while a two-stage solution strategy combining branch-and-bound with genetic algorithms balances accuracy and efficiency. Ultimately, through case validation at a major Chinese shipbuilding enterprise, the proposed task package segmentation methodology demonstrated both its efficacy and practical feasibility. Empirical results demonstrate that this approach offers significant advantages in practical application. It enhances the cohesion of similar processes within task packages, effectively reduces the frequency of cross-package task coordination, and reasonably controls fluctuations in critical resource utilisation. Furthermore, it adapts to the complex scenarios of multi-disciplinary and multi-resource shipbuilding, providing a scientific decision-making tool for project task organisation and resource allocation.