Abstract
This paper proposes and develops a digital twin-based method for assembly deviation analysis, which considers the impact of product shape deviations and uses assembly data in the product design phase to enhance the performance of assembly deviation analysis. First, the overall product assembly deviation analysis process based on twin information is studied. Second, a mating algorithm, which evaluates multi-source assembly information, is proposed and developed. Third, the interaction between adjacent mating feature deviation propagation and an updated iterative mechanism is examined. Finally, a digital twin assembly deviation analysis model based on the new proposed Jacobian-twin approach is established, and validation is conducted through a case study.