基于数字孪生的装配误差建模与溯源分析方法

Aiming at the problems of untimely assembly quality control due to poor timeliness and low accuracy of assembly error traceability analysis, an assembly error modeling and traceability analysis method based on digital twin is proposed. Firstly, a digital twin framework for assembly error analysis is built by the digital twin technology. On this basis, the assembly physical layer, data perception model, geometric model, and equipment motion model are constructed to provide support for subsequent assembly error transfer model construction and traceability analysis. Secondly, an adaptive growth error transfer network is constructed by taking the component features, quality features, and connection relationship among the features extracted from the assembly process information respectively as the nodes and connection relationship among the nodes in the error transfer network, and then an adaptive growth error transfer model is generated by applying the assembly site data obtained from the data model to the weight assignment of the error transfer network. Then, the multi-algorithm dynamic fusion sorting method is introduced to identify the key assembly nodes of the adaptive growth error transfer model, which provides a basis for assembly quality control. Finally, the inverse backtracking algorithm and error transfer intensity algorithm are combined to calculate the error transfer intensity of each node on the error transfer path and to realize the assembly error traceability analysis. The proposed method is used to conduct error traceability analysis experiments on the assembly experimental platform. The experimental results show that the timeliness using the proposed method could be improved by about 40% compared with the traditional offline traceability analysis method under the condition that the probability of quality problems in each assembly process is equal, which indicates that the proposed method can effectively improve the timeliness of assembly quality control and reduce the cost thereof.