Hybrid Mechanism and Data-Driven Digital Twin Model for Assembly Quality Traceability and Optimization of Complex Products

Chao Zhang,Yongrui Yu,Guanghui Zhou,Junjie Hu,Ying Zhang,Dongxu Ma,Wei Cheng,Songchen Men
DOI: https://doi.org/10.1016/j.aei.2024.102707
IF: 8.8
2024-01-01
Advanced Engineering Informatics
Abstract:The digital twin technology has been regarded as one of the vital means to ensure assembly quality and consistency in smart assembly paradigm. However, the mechanism of the coupling effect of multiple assembly parameters on product quality is still unclear, leading to the frequent occurrence of out-of-tolerance. Consequently, a novel hybrid mechanism and data-driven digital twin model (HyDT) is proposed for assembly quality traceability and optimization of complex products. HyDT could firstly perceive the potential assembly quality problem through a forward visual simulation process based on a data-driven model, then identify the specific assembly processes and parameters associated with that problem through reverse root-cause analysis based on a time-series snapshot network-enabled mechanism model, and finally optimize and adjust the associated parameters to ensure the assembly quality and consistency. A HyDT prototype system is thus implemented and demonstrates the feasibility and effectiveness of the proposed approach. Take nozzle assembly as an example, the proposed HyDT could predict the assembly tolerance with high fidelity and improve the assembly quality by an average of 65.61%.
What problem does this paper attempt to address?