A reliability evaluation method for RV reducer by combining multi-fidelity model and Bayesian updating technology

Abstract

RV (rotating vector) reducer is a two-stage transmission mechanism composed of involute planetary gear transmission and cycloid gear transmission with small tooth difference. It is widely used in the transmission mechanism of industrial robots and CNC machine tools. Nonlinear dynamics of gear system is always an important and difficult problem, the randomness of manufacturing and assembly makes the dynamic response become more complex. In this paper, the stochastic dynamic response analysis and dynamic reliability estimation method is proposed for RV reducer. In this method, firstly, considering the uncertainties of manufacturing error and assembly error, the dynamic transmission performance is analyzed respectively using physical prototype test, virtual prototype simulation and theoretical model. Secondly, multifidly model and Bayesian theory are combined to estimate the statistical distribution and distribution parameters of the dynamic transmission error. Finally, the reliability of RV reducer is estimated, and the effects of uncertainties from manufacturing error and assembly error on the transmission accuracy are discussed.