基于改进遗传算法的铸造生产线数字孪生仿真优化  

Digital twin simulation optimization of casting production line based on improved genetic algorithm

作　　者：金秋 原博文 王清岩 JIN Qiu;YUAN Bowen;WANG Qingyan(School of Economics and Management,Tianjin University of Science and Technology,Tianjin 300457)

机构地区：[1]天津科技大学经济与管理学院,天津300457

出　　处：《机械设计》2025年第3期71-77,共7页Journal of Machine Design

基　　金：天津市教育委员会社会科学重大项目(2024JWZD32)。

摘　　要：为了提升铸造生产线性能,提出了一种基于改进遗传算法的铸造生产线数字孪生仿真优化方法。针对遗传算法常面临的固定变异率及交叉率的局限、算法易陷入局部最优、难以实现全局搜索等问题,设计了一种双种群遗传算法。该算法允许在两个不同的种群之间进行信息交换,并通过引入非均匀变异率和交叉率策略,动态调整这些参数以适应优化过程中的不同阶段,从而提升解的质量。文中以某企业铸造生产线为研究对象,根据实际生产系统的布局及工艺流程,构建了铸造生产线的数字孪生模型,采用改进的双种群遗传算法对生产线生产过程进行评估和优化,并将其与传统遗传算法进行对比。结果表明:将数字孪生技术与改进遗传算法相结合,生产平衡率显著提升,生产平滑指数大幅减小,同时算法的收敛速率及结果均得到明显改善。In this article,in order to improve the casting production line’s performance,a method of digital twin simulation optimization is proposed for the casting production line,based on the improved genetic algorithm.The traditional genetic algorithms often suffer limitation in terms of fixed mutation rate and crossover rate;as a result,they tend to get trapped in local optima,making it difficult to achieve global search.In contrast to traditional genetic algorithms,a dual-population genetic algorithm is designed.This algorithm allows information exchange between two different populations,and introduces the strategy of non-uniform mutation rate and crossover rate.On this basis,these parameters are adjusted dynamically,so as to adapt to different stages in the optimization process and ensure better quality of these solutions.With focus on the casting production line of an enterprise,a digital twin model of the casting production line is set up according to the layout and process flow of the actual production system;the improved double-population genetic algorithm is used to evaluate and optimize the production process;this algorithm is compared with traditional genetic algorithms.The results show that a combination of the digital twin technology and the improved genetic algorithm significantly improves the production balance rate and greatly reduces the production smoothness index;the algorithm has a higher standard of convergence speed and effect.

关 键 词：数字孪生 仿真优化 双种群遗传算法 非均匀变异 

分 类 号：TH165[机械工程—机械制造及自动化]