A fidelity evaluation method for digital twin model of aero-engine assembly characteristics

Abstract

Assembly Characteristics Digital Twin Model (ACDTM) enables real-time updates of assembly characteristics in the virtual space, which is crucial for the real-time control of these characteristics. However, the selection of an appropriate model remains a challenge due to the lack of a clear criterion. To address this issue, a fidelity evaluation method is proposed to evaluate accuracy and consistency of ACDTM. This study first analyzes the three dimensions evaluating fidelity and the transmission evolution properties brought by digital twins. Based on this analysis, the fidelity transmission and evolution model are constructed. By combining fidelity network construction methods, Prediction Model (PM) evaluation methods, and input data evaluation methods, a two-level loop evaluation process is proposed. The feasibility of this method is applied in the concentricity of casing assembly. By implementing quantitative fidelity evaluation of ACDTM, this work provides a scientific basis for selecting and optimizing digital twin models in aero-engine assembly.