基于多线程并发的两目标演化算法研究  

Research on two-objective evolutionary algorithm based on multi-thread concurrency

作　　者：史晓艳 张桂珠[1] Shi Xiaoyan;Zhang Guizhu(School of IoT Engineering, Jiangnan University, Wuxi 214122, China)

机构地区：[1]江南大学物联网工程学院

出　　处：《江苏科技信息》2019年第21期28-31,共4页Jiangsu Science and Technology Information

基　　金：国家自然科学基金面上项目;项目名称：基于合作协同演化的大规模黑箱优化的关键问题研究;项目编号：61673194

摘　　要：文章针对基于两目标优化的演化算法计算规模大,消耗大量的计算时间问题,提出了一种基于多线程并发的多目标优化算法。通过引入多线程并发执行框架Executor,实现多核环境下的多任务并发执行,以提高算法的运行效率。该解决方案对多目标演化算法进行重构,使之适用于多线程环境,为进一步优化算法作铺垫。Aiming at the the large scale of two-objective evolutionary algorithm which consumes a lot of computation time, a new algorithm based on multi-threaded concurrency is proposed. On the condition of multi-core environment, the algorithm uses the multi- thread concurrent execution framework executor to achieve multi- task concurrent execution and improves operational efficiency. The solution refactors the multi- objective evolutionary algorithm to make it suitable for multi-threaded environments, and prepares for further optimization of the algorithm.

关 键 词：多目标演化 线程 并发 执行框架 

分 类 号：G20[文化科学—传播学]