Abstract: The modelling of type-A tanks of a liquefied gas carrier is complicated and highly coupled with the principal dimension, hull lines and target cabin capacity. A parametric modelling based subdivision optimization design method for the type-A tank of a liquefied gas carrier (Parametric Optimization Method for Subdivision of type-A tanks, POMSA) has been proposed to improve the efficiency of the overall optimization design of the type-A tank of a liquefied gas carrier. The key technology of this method is the parametric modeling based on the NAPA platform.The typical midsection shape, the longitudinal position of the main transverse bulkhead and the angle of the inner hull are parametrically modelled and regarded as design variables in this paper.The principal dimension, hull lines and the initial general arrangement are taken as input. The requirements of the tank location from IGC, the minimum side length of the tank from the construction process, the target cabin capacity, the propeller immersion and the vision requirement of the ballast water are used as constraints for POMSA. The maximum cabin capacity is taken as the optimization design objective. The optimization model is solved by the optimization mathematical modelling of an exploratory global optimization algorithm (Non-dominated Sorting Genetic Algorithm-II, NSGA-II) together with a local fast convergence optimization algorithm (Hooke-Jeeves direct search method). It is proved that the optimization iteration of this method is efficient, which will facilitate the fast response at the early design stage of the overall design.

Key words: very large liquefied gas carrier (VLGC), type-A tank, subdivision optimization design, parametric modeling, genetic algorithms (GA)

摘要： 液化气船A型舱建模复杂,且与主尺度、线型及舱容指标等耦合性很强,为提高A型货舱的液化气船在总体层面综合优化设计效率,提出了一种基于参数化建模的液化气船A型货舱的分舱优化设计方法（POMSA）。该方法以参数化建模为核心,以NAPA三维设计软件为平台,对液化气船A型货舱包括典型中横剖面形式、主横舱壁位置以及内壳折角三个方面进行参数化建模表达并作为主要设计变量;基于主尺度、线型、初步总布置等信息作为前提输入;以IGC规范对货罐位置要求、建造工艺对货罐最小边长的要求,目标货舱舱容,螺旋桨浸没以及视线等对压载水量的要求为约束;以货舱舱容最大化为优化设计目标;以iSIGHT优化设计软件为平台,搭建优化数学模型,采用一种探索型全局优化算法的组合算法求解优化模型。经验证,该方法优化迭代效率很高,适合总体设计前期快速评估反馈方案。

关键词: 超大型液化气船, A型舱, 分舱优化设计, 参数化建模, 遗传算法