Prediction of random vibration error for double involute gear based on poor information theory
FAN Zhi min, REN Meng cheng
(College of Mechanical Engineering, Qingdao University of Science and Technology, Qingdao 266061, China)

Abstract: Aiming at solving the problems that double involute gear drive has less characteristics data, unknown probability distribution and lack of prior knowledge, based on the theory of poor information, the error prediction of the gear drive was studied in the process of random vibration. Combined with the advantages of GM (1, 1) model and bootstrap sampling principle, a new method was presented based on the dynamic estimation of the poor information, and the dynamic grey bootstrap model was established in the data sample of poor information. The experimental data were used to evaluate the vibration error of the double involute gear, and the predicted values were compared with the original experimental data to verify the reliability of the model, meanwhile compared with the predicted values of GM (1, 1) model and Monte Carlo model to validate the effectiveness of the model. The results indicate that this model has high prediction accuracy, but also can shorten the testing time effectively, reduce the cost of test, and can solve the random vibration error prediction problem of double involute gear preferably. This method can also be used for other gear drive.

Key words: poor information; double involute gear; dynamic grey bootstrap model; vibration error