摘要
针对差分进化易陷入局部最优和灰狼算法易早熟停滞的缺点,提出了一种基于差分进化(DE)算法和灰狼(GWO)算法的混合优化算法(DEGWO)。该算法利用差分进化的变异、选择算子维持种群的多样性,然后引入灰狼算法与差分进化的交叉、选择算子进行全局搜索。在整个寻优过程中,反复迭代渐进收敛。选取此3个测试函数进行仿真验证,结果表明,混合优化算法相比于DE算法和GWO算法,其求解精度、收敛速度、搜索能力都有了显著提高。
In order to overcome these disadvantages that differential evolution is easy to fall into local optimum and grey wolf optimizer behaves premature convergence easily,a hybrid optimization algorithm( DEGWO) based on the combination of differential evolution( DE) and grey wolf optimizer( GWO) is proposed. The differential mutation and selection operations of differential evolution are used to maintain the diversity of the population. Then GWO is introduced to carry out for global exploration,followed by crossover and selection operations. In the whole optimization process,this can be iterated repeatedly and behave convergence gradually. Three test functions were chosen to verify the effect. the results show the hybrid optimization algorithm has significantly improved the accuracy,convergence speed and search ability compared with the DE algorithm and the GWO algorithm.
出处
《科学技术与工程》
北大核心
2017年第16期266-269,共4页
Science Technology and Engineering
基金
吉林省科技发展计划项目(20150203003SF)资助
关键词
差分进化
灰狼算法
混合优化算法
测试函数
differential evolution
grey wolf optimizer
hybrid optimization algorithm
test function