Multi-level assembly deviation prediction of aircraft structure based on grey prediction theory

Due to the small production of aircraft products, the deviation data is small sample sizes, so assembly deviation can't be calculated by the probability statistics method. In addition, because the aircraft structure is flexible, it is easy to deform during assembly, so assembly deviation can't be calculated by assembly dimension chains. Therefore, a multi-level assembly deviation prediction method of aircraft structure is proposed based on the grey prediction theory. Firstly, according to deviation detecting data of assembly part, the deviation prediction model is established. Then the deviation prediction data is calculated based on the deviation prediction model. Secondly, integrating prediction data into the next assembly level, the deviation prediction data of the next assembly level is calculated by building the prediction model of the next assembly level. Finally, the prediction data is fused to the next assembly level continuously, until the assembly deviation prediction data of the product is calculated. The application cases show that the accuracy of the deviation prediction method is 96.0%, which verifies the feasibility and accuracy of the prediction method proposed in this paper.