CAJ | 학술논문

由于传统和声搜索算法每次迭代只生成一个和声向量，因此搜索能力受到限制。针对这一问题，提出一种改进的和声搜索算法。在改进的方法中，每次迭代首先生成与种群规模相同的和声向量，然后放入和声记忆库中并按适应度从高到低排序，最后取前一半个体进入下一代种群。实验结果表明，改进算法不仅在相同迭代步数下明显优于原算法，而且在相同运行时间下也明显优于原算法，从而揭示出该算法能有效地提取种群中的优良个体特征，获得满意的优化结果。
유우전통화성수색산법매차질대지생성일개화성향량，인차수색능력수도한제。침대저일문제，제출일충개진적화성수색산법。재개진적방법중，매차질대수선생성여충군규모상동적화성향량，연후방입화성기억고중병안괄응도종고도저배서，최후취전일반개체진입하일대충군。실험결과표명，개진산법불부재상동질대보수하명현우우원산법，이차재상동운행시간하야명현우우원산법，종이게시출해산법능유효지제취충군중적우량개체특정，획득만의적우화결과。
The only one harmony vector is obtained in each of iteration in classical harmony search algorithm,which affects its search abil-ity. In view of this,propose an improved harmony search algorithm in this paper. In this approach,an equivalent number of harmony vec-tors with population size are obtained in each of iteration,and the newly generated harmony vectors are put into the harmony memory ar-ray. Then,all harmony vectors are sorted according to the fitness from high to low,and the first half individuals are removed into the next generation of populations. Experimental results show that the proposed approach is obviously superior to the classical one under both the same number of iterations and the same running time,which reveals this approach can effectively extract the characteristics of excellent in-dividuals in the population and obtain the satisfactory optimization results.