翻译的科技论文节选
相比遗传算法中的交换和变异操作,当种群的多数个体适应值相差不大时交叉操作就显得无能为力,算法陷入局部解而不能由交换解决,突然变异能够使之摆脱局部收敛而跃出局部解,但是后期的变异可能破坏已产生的对形成最优解有建设性作用的模块。CTOA可以有效避免遗传算法的这个缺点,因为进化式变异和突变均利用了历史搜索结果,优化效果极为显著。
Compared with the exchange and mutation operation in genetic algorithm ,crossover operation becomes helpless when adaptive value of most individules in demes have little difference.The algorithm falls into local solution and can’t be solved by exchange. Although mutation can make it shake off local convergence and jump out of local solution, denteric mutation may destroy the produced module that is constructive to form the optimal solution. CTOA can effectively aviod this shortage of genetic algorithm,since evolving mutation and sudden mutation both make use of historic searching result and optimization effect is very remarkable.