Digital twin-based research on the prediction method for the complex product assembly abnormal events

Abstract

ABSTRACT The emergence of abnormal events (e.g. personnel abnormalities, equipment failures, etc.) on the assembly floor of complex products can seriously affect normal assembly progress. In response to the problems of poor timeliness and lack of predictability in the control of abnormal events on the assembly floor, a method for predicting abnormal events on the assembly floor of complex products based on digital twin technology is proposed. A model for predicting abnormal assembly events is constructed with the physical assembly workshop, virtual assembly workshop, assembly workshop twin data platform and abnormal events prediction service system working together, and its prediction operation mechanism is designed based on the classification of abnormal events and the workflow of the mechanism under different states is analysed. The Grey-Markov method is used to predict abnormal assembly events and provide real-time information for the planning and scheduling system. In order to verify the effectiveness of this scheme, combined with the electrical multiple units bogie assembly workshop, the prediction of the number of equipment failures at the bottleneck station is achieved. The prediction accuracy is much better than that of the GM(1,1) model and can be applied to actual production.