Causal Graph Attention Networks for Quality Prediction of Mass Customization Production Process

The quality prediction in the mass customization production process is a challenge because this manufacturing process is characterized by modular manufacturing and customized assembly, which make traditional quality prediction methods inapplicable. In this paper, a causal graph attention network (CGAT) is proposed to alleviate this problem. A quality prediction framework is proposed for the characteristics of mass customization production process. Then, the mechanism and data are jointly used to construct the causal graph structure. Finally, the graph attention network is used for quality prediction. To verify the effectiveness of the proposed method, The method is tested on the automobile manufacturing simulation data set. The experimental results show that the model has great predictive performance on mass customization datasets.