基于交互式遗传算法的轨道列车造型优化设计研究

Research on optimal design of rail train modeling based on interactive genetic algorithm

面向轨道列车车头造型设计过程中的方案推演与评价调整的低效问题,基于交互式遗传算法开发了设计者与算法协作的交互优化技术。针对轨道列车造型的特征,设计了三线造型编码方法,实现了造型方案的批量化快速生成与优化。通过设计者提供的造型曲线框架,程序可自动完成方案种群的构建,实现造型约束,并可基于设计者的交互评价进行方案换代演化、交叉重组和变异等系列遗传操作。通过草图曲线的智能化识别与调整技术,降低了对设计者手绘输入曲线的精度要求,减轻了设计人员的工作量。设计者与智能技术的合理分工,拓展了概念方案搜索的广度,可提升设计效率。

Aiming at the inefficiency of scheme deduction and evaluation adjustment in the process of rail train’s head modeling design,an interactive optimization technology between designer and algorithm was developed based on interactive genetic algorithm.According to the characteristics of rail train’s modeling,a three-line modeling coding method was designed to realize batch rapid modeling and optimization of modeling schemes.The program could automatically complete the construction of project population and realize modeling constraints through the modeling curve framework provided by designer.A series of genetic operations,such as scheme upgrading evolution,cross-recombination and mutation could be performed based on designer’s interactive evaluation.The accuracy requirements for designers’hand-drawn input and the workload of designers were reduced through the intelligent recognition and adjustment technology of sketch lines.The reasonable division of labor between designers and intelligent technology expanded the breadth of concept solutions searching and improved design efficiency.