基于条件生成模型的船型概念方案正向设计方法探索

doi:10.11996/JG.j.2095-302X.2025061209

图学学报 ›› 2025, Vol. 46 ›› Issue (6): 1209-1215.DOI: 10.11996/JG.j.2095-302X.2025061209

• 制造产品核心工业软件 • 上一篇下一篇

基于条件生成模型的船型概念方案正向设计方法探索

刘德丰¹^,²(), 陈伟政¹^,²(), 白亚强¹^,², 刘凯¹^,², 王琦¹^,²

¹ 中国船舶科学研究中心，江苏无锡 214082
² 深海技术科学太湖实验室，江苏无锡 214082

收稿日期:2025-07-31 接受日期:2025-11-10 出版日期:2025-12-30 发布日期:2025-12-27
通讯作者:陈伟政(1974-)，男，研究员，博士。主要研究方向为船海装备总体性能数字化。E-mail：chenwz.mail@163.com
第一作者:刘德丰(1995-)，男，工程师，硕士。主要研究方向为人工智能与船舶设计交叉应用。E-mail：defeng@cssrc.com.cn

Exploration of forward design methods for ship conceptual schemes based on conditional generative models

LIU Defeng¹^,²(), CHEN Weizheng¹^,²(), BAI Yaqiang¹^,², LIU Kai¹^,², WANG Qi¹^,²

¹ China Ship Scientific Research Center, Wuxi Jiangsu 214082, China
² Taihu Laboratory of Deepsea Technological Science, Wuxi Jiangsu 214082, China

Received:2025-07-31 Accepted:2025-11-10 Published:2025-12-30 Online:2025-12-27
First author：LIU Defeng (1995-), engineer, master. His main research interests cover cross-application of artificial intelligence and ship design. E-mail：defeng@cssrc.com.cn

摘要/Abstract

摘要：

概念设计是指在初步设计阶段，由原始需求提出到初始方案生成的过程。针对船型设计在该阶段需求变动频繁、母型依赖性较强、多方案并发设计成本较高等问题，立足无近似母型场景，探索并提出了一种以船型为对象的条件生成对抗模型，且将阻力性能作为条件标签，生成满足性能指标需求的船型几何方案。首先，利用参数化建模方法构筑无附体船的轮廓特征曲线，筛选获取用以表征该船型的几何参数，并得到相匹配的三维网格。其次，依据积分理论从三维网格获取该船型的特征面面积及船型系数，利用所开发的性能预报平台自动化接口，构建几何参数与性能表征一致性数据集。最后，在传统生成对抗网络(GAN)基础上，通过多层感知机对以阻力为代表的条件特征进行编码，再与生成器隐藏层中的几何特征进行叠加，以引导学习不同阻力表现下的船型样本分布，并生成符合条件约束的船型几何参数。初步实现了不依托母型信息的、从典型总体性能出发的船型正向设计过程，为需求不确定情况下的概念方案快速生成与迭代优化提供了设计基础。

关键词: 船舶设计, 概念设计, 船型方案, 正向设计, 智能生成, 条件生成模型, 生成对抗网络

Abstract:

Conceptual design refers to the process from the proposal of original requirements to the generation of initial schemes during the preliminary design phase. In response to frequent changes in requirements for ship design at this stage, strong dependence on parent models, and the high cost of concurrent design of multiple schemes, a conditional generative adversarial model for hull forms was proposed for scenarios lacking approximate parent models. This model used resistance performance as the conditional label to generate hull-form geometric schemes that met the performance index requirements. Firstly, contour feature curves of the bare hull were constructed using parametric modeling methods; geometric parameters characterizing the hull form were selected, and corresponding three-dimensional meshes were obtained. Secondly, based on the integral theory, the characteristic surface areas and hull-form coefficients were obtained from the three-dimensional mesh, and a consistency dataset of geometric parameters and performance characteristics was constructed via the automated interface of the developed performance-prediction platform. Finally, based on a traditional generative adversarial network (GAN), a multilayer perceptron was employed to encode conditional features represented by resistance, which were then combined with geometric features in the hidden layers of the generator. This approach guided the learning of hull-form sample distributions under varying resistance conditions and generated hull-form geometric parameters that satisfied the specified constraints. The initial realization of a forward-design process for hull forms without reliance on parent-model information and based on overall performance of typical types provided a design basis for the rapid generation and iterative optimization of conceptual schemes under uncertain requirements.

Key words: ship design, conceptual design, hull design concept, forward design, intelligent generation, conditional generation model, generative adversarial networks

中图分类号:

U662

刘德丰, 陈伟政, 白亚强, 刘凯, 王琦. 基于条件生成模型的船型概念方案正向设计方法探索[J]. 图学学报, 2025, 46(6): 1209-1215.

LIU Defeng, CHEN Weizheng, BAI Yaqiang, LIU Kai, WANG Qi. Exploration of forward design methods for ship conceptual schemes based on conditional generative models[J]. Journal of Graphics, 2025, 46(6): 1209-1215.

图/表 5

图1 部分船型几何参数修改对三维建模的影响

Fig. 1 The influence of partial modification of hull geometry parameters on 3D modeling

图2 船型数据集的生成过程

Fig. 2 The generation process of ship datasets

图3 所使用的条件生成模型

Fig. 3 The conditional generation model used

图4 船型生成结果及典型尺度/船型参数信息((a) 生成结果1；(b) 生成结果2)

Fig. 4 Generation results of ship hull and information of scale/ship form ((a) The generated result 1; (b) The generated result 2)

图5 生成的船型方案在15节航速下阻力表现情况((a) 中横剖面图；(b) 侧面图)

Fig. 5 The resistance performance of the generated ship hull schemes at a speed of 15 knots ((a) Midship section; (b) Profile view)

参考文献 21

[1]	XIAO G, YANG D, XU L, et al. The application of artificial intelligence technology in shipping: a bibliometric review[J]. Journal of Marine Science and Engineering, 2024, 12(4): 624. DOI URL
[2]	FAHRNHOLZ S F, CAPRACE J D. A machine learning approach to improve sailboat resistance prediction[J]. Ocean Engineering, 2022, 257: 111642. DOI URL
[3]	LANG X, WU D, MAO W G. Comparison of supervised machine learning methods to predict ship propulsion power at sea[J]. Ocean Engineering, 2022, 245: 110387. DOI URL
[4]	SHEN Y J, YE S X, ZHANG Y W, et al. Application of machine learning for bulbous bow optimization design and ship resistance prediction[J]. Applied Sciences, 2025, 15(6): 2934. DOI URL
[5]	GUPTA P, RASHEED A, STEEN S. Ship performance monitoring using machine-learning[J]. Ocean Engineering, 2022, 254: 111094. DOI URL
[6]	LEVINE M D, EDWARDS S J, HOWARD D, et al. Multi-fidelity data-adaptive autonomous seakeeping[J]. Ocean Engineering, 2024, 292: 116322. DOI URL
[7]	LIU J, ZHANG B J, HU L F, et al. Research on hull form optimization at multiple speeds based on machine learning and ship model experiments[J]. Engineering Applications of Artificial Intelligence, 2025, 160: 111882. DOI URL
[8]	BAGHERI H, GHASSEMI H. Genetic algorithm applied to optimization of the ship hull form with respect to seakeeping performance[J]. Transactions of FAMENA, 2014, 38(3): 45-58.
[9]	ZHANG S L, CHEN Y Z. Research on deep learning method in ship hull form optimization[C]// The 3rd International Conference on Computer Science and Communication Technology. Beijing: SPIE, 2022: 125066O.
[10]	WEI Y B, WANG Y J, WAN D C. Hull form optimization based on multi-fidelity deep neural network[J]. Chinese Journal of Ship Research, 2024, 19(6): 74-81.
[11]	MIRJALILI S, SAREMI S, MIRJALILI S M, et al. Multi-objective grey wolf optimizer: a novel algorithm for multi-criterion optimization[J]. Expert Systems with Applications, 2016, 47: 106-119. DOI URL
[12]	YONEKURA K, OMORI K, QI X, et al. Designing ship hull forms using generative adversarial networks[J]. AI, 2025, 6(6): 129. DOI URL
[13]	GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[J]. Communications of the ACM, 2020, 63(11): 139-144. DOI URL
[14]	杜林, 李胜忠, 李广年, 等. 基于深度卷积生成式对抗网络的船型特征认知与条件生成方法[J]. 船舶力学, 2024, 28(8): 1162-1174.
	DU L, LI S Z, LI G N, et al. A ship hull offset feature cognition and generation method based on conditional deep convolutional generative adversarial networks[J]. Journal of Ship Mechanics, 2024, 28(8): 1162-1174 (in Chinese).
[15]	THAKUR S, SAXENA N V, ROY P S. Generative AI in ship design[EB/OL]. [2025-03-31]. https://arxiv.org/pdf/2408.16798.pdf.
[16]	KERAMAT H, KIRCHEN P, HANNAN M, et al. A reward-directed diffusion framework for generative design optimization[EB/OL]. [2025-03-31]. https://arxiv.org/abs/2508.01509v1.pdf.
[17]	KIM J H, ROH M I, YEO I C. Hull form optimization of fully parameterized small ships using characteristic curves and deep neural networks[J]. International Journal of Naval Architecture and Ocean Engineering, 2024, 16: 100596. DOI URL
[18]	BAGAZINSKI N J, AHMED F. ShipGen: a diffusion model for parametric ship hull generation with multiple objectives and constraints[J]. Journal of Marine Science and Engineering, 2023, 11(12): 2215. DOI URL
[19]	BAGAZINSKI N J, AHMED F. C-ShipGen: a conditional guided diffusion model for parametric ship hull design[EB/OL]. [2025-03-31]. https://arxiv.org/pdf/2407.03333.pdf.
[20]	KHAN S, GOUCHER-LAMBERT K, KOSTAS K, et al. ShipHullGAN: a generic parametric modeller for ship hull design using deep convolutional generative model[J]. Computer Methods in Applied Mechanics and Engineering, 2023, 411: 116051. DOI URL
[21]	BAGAZINSKI N J, AHMED F. Ship-D: ship hull dataset for design optimization using machine learning[EB/OL]. [2025-03-31]. https://arxiv.org/pdf/2305.08279.pdf.

基于条件生成模型的船型概念方案正向设计方法探索

Exploration of forward design methods for ship conceptual schemes based on conditional generative models

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 5

参考文献 21

相关文章 15

编辑推荐

Metrics

本文评价

[1]	刘宗明, 洪唯, 龙睿, 祝越, 张小宇. 基于自注意机制的乳源瑶绣自动生成与应用研究[J]. 图学学报, 2024, 45(5): 1096-1105.
[2]	朱宝旭, 刘漫丹, 张雯婷, 谢立志. 高分辨率人脸纹理图全流程生成方法[J]. 图学学报, 2024, 45(4): 814-826.
[3]	贺文虎, 刘伟, 王酉龙, 蔡珲. 航空动力装备顶层功能分解分配建模方法研究[J]. 图学学报, 2024, 45(2): 292-299.
[4]	古天骏, 熊苏雅, 林晓. 基于SASGAN的戏剧脸谱多样化生成[J]. 图学学报, 2024, 45(1): 102-111.
[5]	徐祯东, 张天宇, 张世恒, 姚从荣, 王道累. 基于YUV颜色空间GAN网络的图像去雾算法研究[J]. 图学学报, 2023, 44(5): 928-936.
[6]	邵俊棋, 钱文华, 徐启豪. 基于条件残差生成对抗网络的风景图生成[J]. 图学学报, 2023, 44(4): 710-717.
[7]	廖仕敏, 刘仰川, 朱叶晨, 王艳玲, 高欣 . 一种基于 CycleGAN 改进的低剂量 CT 图像增强网络[J]. 图学学报, 2022, 43(4): 570-578.
[8]	方洪波, 万广, 陈忠辉, 黄以卫, 张文勇, 谢本亮. 基于改进 YOLOv5s 的离线手写数学符号识别[J]. 图学学报, 2022, 43(3): 387-395.
[9]	汪玉金, 谢诚, 余蓓蓓, 向鸿鑫, 柳青. 属性语义与图谱语义融合增强的零次学习图像识别[J]. 图学学报, 2021, 42(6): 899-907.
[10]	林森 , 刘旭 . 门控融合对抗网络的水下图像增强 [J]. 图学学报, 2021, 42(6): 948-956.
[11]	杨勇，刘惠义. 极端低光情况下的图像增强方法[J]. 图学学报, 2020, 41(4): 520-528.
[12]	李桂，李腾. 基于姿态引导的场景保留人物视频生成[J]. 图学学报, 2020, 41(4): 539-547.
[13]	陶桂林1，马文玉 1，唐克强 2，杜奕呈 3. BIM 正向设计存在的问题和思考[J]. 图学学报, 2020, 41(4): 614-623.
[14]	李天赠，黄红梅，陈家静，赖春敏. 基于实验分析与数值模拟技术的产品概念设计方法[J]. 图学学报, 2020, 41(1): 132-140.
[15]	罗琪彬 1,2，蔡强 1,2 . 采用双框架生成对抗网络的图像运动模糊盲去除[J]. 图学学报, 2019, 40(6): 1056-1063.