混合正交遗传算法及其在函数优化上的应用

在分析和研究正交遗传算法的基础之上,依据混合优化策略及混合遗传算法的构造原则,通过对自适应正交局部搜索算子的改进提出了一种新的变异算子。该算子具备自适应全局搜索和局部搜索的能力,能够保证算法的变异概率取值为1.0时,算法的搜索效率最高;结合正交交叉算子之后,又能保证算法的交叉概率也取值为1.0时,算法的搜索效率最高;由此解决了交叉概率和变异概率参数的匹配问题。而使用的截断选择和负相关配对、最优交叉策略、精英选择和重复个体剔除策略等组合算子,一方面能够保证算法的收敛速度;另一方面也能有效地保持种群的多样性,这样在保证算法快速收敛的同时避免出现早熟现象;由此解决了＂全局最优＂和＂快速收敛＂的矛盾。因此,提出的改进型新算法在处理一些常用的测试函数上具有较高的效率。

Hybrid Orthogonal Genetic Algorithm and Its Application in Function Optimization

According to the hybrid optimization strategy and the rules of hybrid genetic algorithms,it creates a new mutating operator — orthogonal mutation operator on the basis of the operator of adaptive orthogonal local search.This kind of operator can provide with adaptive global search and local search at the same time.It can keep the most effective searching when the probability of mutation means 1.0,combining using the orthogonal crossover operator;it can keep the most effective searching when the probability of crossover means 1.0.And then,it solves the problem of the optimizing to the group of uncertain probabilities of crossover and mutation in genetic algorithm.Moreover,the paper has designed some groups of new operators such as truncation selection and negative-correlation partnership,best-fit-crossover strategy,the best individual holding strategy and repeated individual elimination,etc.On the one hand,these hybrid operators can keep the speed of searching,on the other hand,they can provide global search（reliability） by means of the promotion of high levels of population diversity.And then,it solves the problem of the contradiction between the ＂high speed constringency＂ and ＂the most global-optimized＂.As a result,the method is effective in the dealing with some common function.