考虑重合度的斜齿轮传动优化混合遗传算法

Hybrid Genetic Algorithm for Optimal Design of Helical Gear Transmission Considering Coincidence Degree

以体积最小和重合度最大为目标函数,建立了斜齿轮传动多目标优化设计数学模型,并将问题转化为无约束单目标优化问题。针对优化设计参数的特点,遗传算法的编码方式采用整数编码和实数编码相结合的混合编码。操作过程中,结合了模拟退火算法调整适应函数,设计了多父辈随机交叉方法,改进了变异操作,从而形成了改进的混合遗传算法。优化过程中,通过编码及操作方法的设计部分约束条件自动得到满足,减少了不可行解的产生。算例说明了该优化方法的有效性。

A mathematical model of the optimal design of helical gear transmission is established, in which the least volume and the maximum coincidence degree are taken as objective function. Aiming at the feature of the design parameter, the hybrid coding combining integer coding with real coding is adopted. In the operational process, simulated annealing algorithm is combined to adjust the fitness function, multi-parent random cross method is designed, and mutation method is improved, then the improved hybrid genetic algorithm is formed. In the optimization procedure, through the designs of coding and operational method, some restrictions are satisfied automatically, so the number of infeasible solutions is reduced. Example shows that the algorithm is efficient.