A new approach to the hull-form design of inland cargo ship based on knowledge graphs

Currently, the complex navigation environment of the Yangtze River, the diversity of ship types, and the high energy consumption and pollution levels of some vessels create an urgent need for energy-efficient and intelligent inland ship designs. Traditional design methods often cannot satisfy the requirements of rapid hull-form optimization. This paper constructs and applies a hull-form knowledge graph for inland cargo ships. A parametric hull model is established in CAESES, and RANS-based CFD simulations are integrated within a simulation-based design optimization (SBDO) framework to generate the resistance database for sensitivity analysis, surrogate-model training, and knowledge-graph construction. The CFD model is validated against towing-tank tests and is further used to verify hull forms generated by the proposed knowledge-graph interpolation method. Classification principles, a construction process, and an application process are proposed. Two application cases demonstrate that the method can rapidly generate hull forms with resistance performance comparable to or better than conventional parent-ship designs.