Reliability design and optimization of the planetary gear by a GA based on the DEM and Kriging model

Da Cui,Guoqiang Wang,Yanpeng Lu,Kangkang Sun

DOI: https://doi.org/10.1016/j.ress.2020.107074

2020-11-01

Abstract:Based on stress-strength interference theory, calculating the reliability by using the coefficient of variation is a common method for the optimal design of a planetary transmission system. This method calculates the reliability by assuming the tangential force is zero while this is quite different from the real working condition. In this paper, the authors propose a novel reliability design and optimization method of the planetary gear, using the genetic algorithm, based on Kriging model. The Kriging method is used to establish the gear reliability model to simplify the reliability calculation. Then the Kriging model is optimized by using the genetic algorithm to ensure the global optimal solution. To simulate the real working conditions of planetary gears, the discrete element method (DEM) is adopted to calculates the load variation coefficient of the planetary gear. By taking the double toothed roller crusher as case study, the optimization results show that proposed method can significantly improve the calculation efficiency, and compared with the traditional design, the volume increases by 36.96%, and the failure rate of the planetary gear decreases by 17.05%.

engineering, industrial,operations research & management science