结合形状文法与交互式遗传算法的列车头型设计
戚百灵1孙元2闫宇婷2林凤玲2
Train Vehicle Nose Design Combining Shape Grammar and Interactive Genetic Algorithm
QI Bailing1SUN Yuan2YAN Yuting2LIN Fengling2
-
作者信息:1.中车大连机车车辆有限公司,116022,大连
2.大连理工大学建筑与艺术学院,116023,大连
-
Affiliation:1.CRRC Dalian Co., Ltd., 116022, Dalian, China
2.School of Architecture and Art, Dalian University of Technology, 116023, Dalian, China
-
关键词:
-
Key words:
-
DOI:10.16037/j.1007-869x.2024.10.011
-
中图分类号/CLCN:TP302； U270.3
-
栏目/Col:研究报告
摘要:
［目的］典型的列车头型美学设计方法普遍依赖于设计人员的主观经验和直觉，已不能满足快速迭代和大规模定制的设计与生产需求。因此，针对列车头型的美学设计问题，提出了一种结合形状文法与交互式遗传算法的列车头型生成式设计方法。［方法］对列车头型不同层次的形状特征线进行分析与提取，结合形状文法对设计元素特征线进行抽象创作；采用交互式遗传算法对造型设计特征进行编码，考虑到用户在评价方案过程中审美认知的模糊性，适应度采用模糊赋值方法，实现头型设计的智能多样化选择；以磁悬浮列车头型设计为例进行验证。［结果及结论］与传统交互式遗传算法的精确赋值相比，采用高斯隶属函数来模糊表示进化个体的适应度值，可以降低用户评价过程中的心理负担。结合形状文法与交互式遗传算法的列车头型生成式设计方法，可以有效改善概念方案质量，提升设计效率。在交互评价过程中，随着进化代数的增加，用户针对方案的平均交互评价时间逐渐减少，这体现了用户对方案理解的渐进性，也说明了交互式生成技术方法符合用户的认知模式。
Abstracts:
［Objective］ The typical train nose aesthetic design method generally relies on the subjective experience and intuition of designers, and can no longer meet the design and production needs of rapid iteration and mass customization. Therefore, aiming at the aesthetic design problem of train nose, a generative design method of which combining shape grammar and interactive genetic algorithm is proposed. ［Method］ The shape feature lines at different levels of the train nose are analyzed and extracted, and the feature lines of the design elements are abstractly created in combination with shape grammar. An interactive genetic algorithm is used to encode design features. Considering the ambiguity of users' aesthetic cognition in the scheme evaluation process, a fuzzy assignment method is used for the fitness to achieve intelligent and diversified selection of the vehicle nose design. The design of maglev train nose is used as an example for verification. ［Result & Conclusion］ Compared with the accurate assignment of traditional interactive genetic algorithms, using Gaussian membership functions for fuzzy representation of the fitness values of evolving individuals can reduce the psychological burden of users in the evaluation process. The train nose generative design method combining shape grammar and interactive genetic algorithm can effectively improve conceptual scheme quality and enhance design efficiency. In the interactive evaluation process, with the increase of evolutionary algebra, the average user's interactive evaluation time for the scheme gradually decreases, which reflects the gradual understanding of the scheme by users, and also shows that the interactive generation technology method conforms to the user's cognitive mode.
- 上一篇: 高铁物流运输模式及其可行性
- 下一篇: 规划引领，立足实际，助力城市轨道交通高质量发展
过刊浏览
资料下载
友情链接