Aiming at the problem that it is difficult to accurately predict the assembly accuracy of aeroengine rotor after considering the assembly deformation of rotor parts

a method for predicting flexible assembly accuracy of aeroengine rotors based on mechanism guidance data was proposed. Firstly

the rigid assembly model of aeroengine rotors was established based on the error transfer and accumulation mechanism. Then

by analyzing the influence of the deformation of mating parts on the assembly accuracy

a flexible assembly accuracy prediction model was proposed. Secondly

combined with the assembly mechanism model of aeroengine rotors

the feature sensitivity of flexible assembly process of rotors was analyzed by numerical analysis method and the feature engineering was established. Thirdly

a prediction model for assembly accuracy of aeroengine rotors based on support vector regression and extreme gradient lift algorithm (SVR-XGBoost) was proposed based on the nonlinear characteristics and uncertainty of aeroengine rotor assembly. Finally

the primary disk and secondary disk of a certain type of aero-rotor were taken as the research object. By comparing the prediction results of different machine learning models

MAE

MSE and R2 evaluation indexes of the machine learning model in this paper are superior to the single SVR and XGBoost methods

and have higher generalization and accuracy. The assembly accuracy prediction results of the model in this paper are in good agreement with the assembly test results

and the average error of the model is reduced by 73.10% compared with the rigid prediction model

which has a significant effect on improving prediction accuracy of the assembly accuracy of the aeroengine rotors.