Large-diameter cabin sections are widely used in the aerospace field and serve as core components of aerospace equipment. Their assembly quality directly affects equipment performance and mission safety. Recently, the Shenyang Institute of Automation (SIA) of the Chinese Academy of Sciences has made progress in the high-precision intelligent assembly of large-diameter cabin sections. The research team innovatively proposed a digital twin-driven virtual-physical collaborative assembly method, which provides a novel technical pathway for the intelligent assembly of heavy-duty aerospace equipment such as solid rocket motors.

The research results were published in the Journal of Manufacturing Systems, under the title Digital twin-based high-precision assembly method for large-diameter cabin section docking. Associate Researcher LIU Mingyang from SIA is the first author, and Researcher XU Zhigang is the corresponding author. Relevant research achievements of the team were selected as one of the Top 10 Scientific and Technological Advances in China's Intelligent Manufacturing 2024.

In the assembly of aerospace equipment, take large-diameter solid rocket motors as an example: during the docking and assembly process of the nozzle and the combustion chamber, the mating features are invisible. To ensure assembly accuracy, repeated manual trial-fitting and calibration are required, making the improvement of assembly efficiency and precision an ever more pressing challenge that urgently demands an efficient and accurate solution.

The team constructed a fully closed-loop intelligent assembly system that spans from 3D scanning and multi-camera online tracking to virtual perception and physical execution.

They developed a feature-marked digital twin modeling method, which transforms the complex and invisible spigot features into visually trackable target features, thereby enabling online measurement and virtual perception of the mating features. They also optimized a pose estimation strategy based on multiple feature points, effectively improving the accuracy of visual pose estimation. The key pose measurement accuracy is better than 0.05 mm, offering reliable assurance for the operational quality and efficiency of the precision assembly process.

Through virtual sensors, the system realizes online monitoring of the docking gap and collision warnings, enabling simulation verification before assembly, dynamic deviation correction during assembly, and risk prediction throughout the entire process. Experimental validation shows that this method achieves a first-attempt docking success rate of ≥95%, which can reduce the need for repeated trial-fitting, thus enhancing assembly efficiency and safety.