Optimization of metal belt continuously variable transmission pulley based on multi-island genetic algorithm

The efficiency of metal belt CVT is an important index to evaluate its transmission performance, and the greater the axial deformation of the pulley, the greater the efficiency loss. In this paper, the deformation of pulley under different passive pressures is measured by experiment, and the reliability of the optimized 3D CAD model is verified by comparing with the data of the simulation model. And based on multi Island genetic algorithm (MIGA), using ISIGHT to integrate CATIA and ABAQUS, and the maximum stress is used as the constraint condition, penalty function integrate pulley deformation, mass and moment of inertia of the smallest to construct single objective function to optimize a CAD model, finally the test sample analysis verify the reliability of the optimization results. Research results show that, the optimized pulley mass was reduced by 16.96% and the inertia was reduced by 11.7%. Comparing the axial deformation of different speed ratios of the pulley, the deformation is smaller than that before the optimization, laid the foundation for the study of pulley transmission efficiency. (paper)