基于自适应正交局部搜索算子的混合遗传算法

The hybridized genetic algorithm based on the operator of adaptive orthogonal local search

基于遗传算法的动态特性和正交设计的思想 ,提出了能根据当前进化的种群状态自适应调整局部搜索空间大小的正交局部搜索算子 .对结合了自适应正交局部搜索算子的混合遗传算法进行了经典的多峰值测试函数的性能测试 ,结果表明 ,混合算法在获得的解的准确性和收敛速度上均优于标准遗传算法 .

On the basis of the theory of Orthogonal Design and the dynamic attribute of GA,this paper presented the operator of orthogonal local search. It can vary the size of the local search area adaptively in response to the current state of the population during the evolution. The hybridized GA combined with the proposed AOLS operator was evaluated on a set of benchmark functions. Results showed that the hybridized algorithm outperformed the traditional GA in terms of solution accuracy and convergence speed.