交叉概率对遗传算法运算速度影响的测试与分析

The Testing And Analysis of the Effects of Crossover Probability on Genetic Algorithms' Calculating Speed

在现有文献基础上,通过理论分析,提出了交叉概率应取1.通过4个测试函数的测试结果,表明交叉概率越大,运算速度越快,平均运行代数越少,验证了通过理论分析提出的交叉概率应取1的科学性和可行性.同时克服了交叉概率难以确定以及确定交叉概率时缺少理论依据的不足.

We proposed the crossover probability should be 1 through theoretical analysis on the basis of existing literature. The results from the trial calculation of the four test functions shows that as the crossover probability increases,the faster the calculation speed and the less the number of average iterations.and they verify the scientificalness and feasibility of which the crossover probability value shoule be 1 proposed by theoretical analysis.In the meantime^the strategy can overcome the shortage of traditional genetic algorithm whose crossover probability value is difficult to determine,and the lack of theoretical basis when we determine the crossover probabilty value.