基于装配语义与图神经网络的产品虚拟装配线智能化构建方法

Intelligent Construction Method of Virtual Assembly Lines for Products Based on Assembly Semantics and Graph Neural Networks

随着智能制造技术的快速发展,数字化工厂方法在生产系统规划设计中的应用日益广泛。构建虚拟装配线或装配线数字孪生体已成为实现装配线数字化规划设计的关键环节。然而,传统的设计方法主要依赖于经验,存在效率低下、设计周期长、优化困难等问题。为了克服上述困难,本研究以产品装配线设计为背景,提出一种基于产品装配语义和图神经网络的产品虚拟装配线智能化构建方法。为验证所提出方法的可行性和有效性,本研究以某液压缸虚拟装配线构建为案例进行了实证研究。结果表明,该方法能够提高装配线设计效率,缩短设计周期,并实现装配线布局的智能优化。

With the rapid development of intelligent manufacturing technology,digital factory methods are increasingly used in production system planning and design.Constructing virtual assembly lines or digital twins of assembly lines has become a key link in realizing the digital planning and design of assembly lines.However,traditional design methods mainly rely on three-dimensional models and manual experience,which suffer from issues such as inefficiency,long design cycle,and difficulties in optimization.To overcome the above challenges,this paper proposes a method for intelligent construction of product virtual assembly line based on product assembly semantics and graph neural networks in the context of product assembly line design.To verify the feasibility and effectiveness of the proposed method,a case study was conducted on the construction of a virtual assembly line for a certain hydraulic cylinder.The results indicate that the method can significantly improve the efficiency of assembly line design,shorten the design cycle,and achieve intelligent optimization of the assembly line layout.