【摘要】 提出了一种基于分流机制的新型遗传算法。该算法采用优种限量繁殖 ,达标种交叉和劣种变异的策略 ,突出遗传算子各自的优点和作用 ,根据进化质量自适应地调整交叉和变异概率 ,使种群具有很强的可进化性。分流机制遗传算法改变了遗传算法的传统结构 ,种群始终具有探测新的超平面 ,从而获得新的最优个体的能力 ;同时 ,对于进化中的个体不需进行所有遗传算子操作 ,提高了算法的全局收敛性 ,降低了时间复杂度。测试结果表明了所提方法的有效性。更多还原

【Abstract】 A new genetic algorithm based on diffluent mechanism is put forward in this paper. Its strategy is that excellent seeds are reproduced in a limited quantity qualified, seeds are crossed and bad seeds are mutated. The virtue and function of genetic operator are developed. The crossover probability and mutation probability are adjusted by the evolution quality. In this way, The seeds acquire very strong capability of evolution. Diffluent mechanism genetic algorithm has changed the traditional structure of genetic algorithm. the seeds can detect new hyperplane all the time, thereby new best individuals are attained. Moreover, it is not necessary that all genetic operators operate for individuals during evolving. The global convergence is improved, the time complexity is depressed in the meantime. The result of simulation demonstrates its validity.更多还原