A prediction model of gear radial composite deviation based on digital twin mesh

The intelligent development of traditional detection technology through the application of emerging technologies is a crucial trend in achieving multi-information gear detection. In this paper, digital twin mesh (DTM) is proposed, and according to the proposed DTM, the prediction model (DTM model) of gear radial composite deviation is established. The DTM model is established by merging the mechanism and the data model with a parallel structure. The mechanism model is established based on the measurement principle. The data model is established through the Gaussian process regression algorithm. Comparing the predicted results of DTM model with the measured results, it is observed that the DTM model demonstrates superior precision. The DTM model achieves a prediction accuracy exceeding 87 % for total radial composite deviation and exceeding 90 % for toothto-tooth radial composite deviation. This DTM model holds significant engineering application value and offers insights for the non-contact dynamic measurement method of gears.