基于混合遗传算法的齿轮传动优化设计

以体积最小为目标函数，建立了齿轮传动优化设计数学模型，并用外部惩罚函数法将该问题转化为无约束优化问题。针对遗传算法的局限性，采用整数编码和实数编码结合的混合编码，并调整了适应函数，采用随机多父辈适应函数值加权交叉和自适应变异操作，结合了模拟退火算法，给出了初温的确定方法，从而形成了混合遗传算法。该算法能够有效地减少不可行解的产生，提高收敛速度，避免早熟收敛。算例说明，该优化方法有效、实用。

Optimal design of gear transmission based on hybrid genetic algorithm

Mathematical model of gear transmission optimization design is built with minimum volume as objective function. The optimization problem is transformed into non - restraint optimization problem using the outer penalty function method. Considering limitation of ordinary genetic algorithm, hybrid genetic algorithm is put forward and used, in which hybrid coding based on integer and real coding is adopted, fitness function is adjusted, random parent - number fitness - weighted crossover and adaptive mutation methods are adopted, simulated annealing algorithm is added to, and method of determining initial temperature is given. The hybrid genetic algorithm features less infeasible solution, higher convergence speed and can avoid premature convergence. Optimization results show that the algorithm is practical and efficient.