【摘要】 针对标准遗传算法(SGA)的随机性大、收敛速度慢等缺点,提出一种通用的改进遗传算法,引入局部补差算子,有效地提高了算法的收敛效率。建立了基于模糊因子函数的适应度函数,使对个体的评价更具有合理性。通过对交叉、变异概率的动态调整,克服了SGA未成熟收敛的弊端。将模拟退火算法与现行终止条件结合,形成了模拟退火收敛准则,极大地改善了SGA的局部搜索能力。据此编制计算程序,将其应用于多约束、多变量、复杂非线性的各类齿轮传动优化设计中,均得到了更好的结果。更多还原

【Abstract】 An improved genetic algorithm in common use is proposed.This algorithm introduces local compensating operator,avoids shortcoming of big randomness and slow convergence of SGA(Standard Genetic Algorithm).(Adaptation) function based on fuzzy factor is set up makes the individual evaluation rational and objective.Through regulating the crossover and mutation rate,it overcomes the disadvantage of premature convergence in SGA.Combining the simulated annealing algorithm with prevail termination condition,a convergence regulation based on simulated annealing algorithm is constructed and the local searching ability of SGA is improved,then the computer program is produced.It is applied to the optimization design of gear transmission with more constraints,variables and complicated non-linear factors,and much better results are obtained.更多还原