Online optimization method for positioning accuracy in cylindrical components aligning based on digital twins

As the production mode of aerospace equipment changes to intelligence, the aligning of cylindrical components increasingly adopts digital twin systems for assembly. However, measurement and transmission processes can affect the accuracy in pose adjustment and positioning. Notably, in the process of automatic aligning, conventional optimization methods fail to compensate for globally accumulated errors in multi-system collaboration, while the absence of real-time monitoring and online optimization further undermines aligning quality and efficiency. To address this challenge, a digital twin-driven online positioning accuracy optimization method is proposed for cylindrical components aligning. In this study, a digital twin system framework for cylindrical components aligning is first established, incorporating a closed-loop control methodology. Subsequently, the method systematically investigates error factors in multi-system coordination, including modeling and analysis of their impacts on positioning accuracy. Key innovations involve an iterative Singular Value Decomposition-based measurement pose optimization algorithm, a mechanism-data fusion-driven online prediction model for alignment errors, and a geometric-analytical cross-system transformation method for precise actuator parameter calculation. By integrating online precision refinement algorithms and predictive error compensation mechanisms, the cumulative positioning errors are effectively mitigated, enhancing dynamic aligning accuracy control capabilities. Experimental validation using a prototype system demonstrates that the method improve the aligning accuracy by 70.77% and shortens the docking cycle time by 53.10%, effectively improving both docking and efficiency, and verifying the correctness and effectiveness of the proposed method.