Digital twin-driven multi-dimensional assembly error modeling and control for complex assembly process in Industry 4.0

Abstract

With the sustainable development of advanced manufacturing technology, the important link affecting the final quality of complex products has been transferred from design and manufacturing to the assembly process. However, traditional pre-defined assembly parameters could not respond to changes in the assembly process caused by assembly errors, leading to unstable assembly quality and poor consistency. To address these issues, this paper proposes a novel online modeling and control method for complex assembly process based on digital twins (DT) and an extended particle swarm optimization algorithm, which could dynamically adjust the assembly process parameters to quickly respond to changes in the assembly process. Firstly, a DT-enabled autonomous assembly framework is proposed for the online analysis, optimization and control of assembly quality, by considering the coupling effect of multi-dimensional assembly errors on assembly quality. Then, two key enabling technologies of the framework, including the multi-dimensional assembly error-fused DT modeling method and DT-driven online optimization and control method of assembly errors are introduced in detail. Finally, a smart assembly prototype is implemented, where its application examples and evaluation results show the superiority of the proposed approach. Compared with traditional method, the proposed approach could significantly reduce the iteration time by 64% and improve the assembly quality by 14.05%.