Real-Time Assembly Guidance System of Aero-Engine Based on Mixed Reality
Mixed reality (MR) provides visualization aids for the assembly of major complex equipment, but there is a lack of research on real-time assembly guidance for aero-engines, so this paper designs a real-time assembly guidance system for aero-engines based on mixed reality. Firstly, the main framework of the system is elaborated from the server side and the client side, in which the 3D digital process manual on the server side is used for the construction of the aero-engine assembly process model, and personalized design can be carried out for different parts; the assembly inspection subsystem is used for the inspection of the assembly results during the assembly process and real-time feedback, and the subsystem can store and call a variety of deep learning models; Secondly, the client side The assembly inspection subsystem is used to inspect the assembly results and provide real-time feedback during the assembly process. Finally, through the real-time assembly guidance test of several aero-engine components, the results show that the guidance system can realize the online guidance assembly of aero-engine components, and the real-time frame rate of the system reaches 28.2, 36.7, and 31.5 FPS, which can meet the real-time demand of the MR assembly guidance process. By analyzing the performance of assembly result detection, the average recognition correctness of this system reaches 93.9%. This provides an intelligent error detection method for MR guided assisted assembly and improves the efficiency of guided assembly. It further proves the effectiveness of the system designed in this paper.
Authors
- Wenjun Zhao (ORCID: https://orcid.org/0000-0001-5368-3181)
- C.H. Chen
- Nansong Wang
- Hao Xu (ORCID: https://orcid.org/0000-0002-7584-8554)
Institutions
- Shanghai Electric Apparatus Research Institute (CN)
Publication Details
- Published
- 2024-10-18
- DOI
- https://doi.org/10.1109/meae62008.2024.11026547
- Primary Topic
- Augmented Reality Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00