融合遗传和粒子群算法的云工作流调度算法  

Workflow scheduling algorithm integrating genetic algorithm in particle swarm optimization algorithm in cloud

作　　者：张宇 ZHANG Yu(Software Institute,Henan Finance University,Zhengzhou 450000,China)

机构地区：[1]河南财政金融学院软件学院,河南郑州450000

出　　处：《计算机工程与设计》2021年第10期2867-2875,共9页Computer Engineering and Design

基　　金：教育部产学合作育人基金项目(201802198003)。

摘　　要：针对云工作流调度问题,提出一种融合遗传算法和粒子群优化算法的工作流调度负载均衡算法。充分利用多元启发式方法融合的优势,避免遗传算法的收敛过慢和粒子群算法易于陷入局部最优的缺陷,有效将工作流任务映射至虚拟机资源,实现全局工作流执行跨度最小化和虚拟机分配的负载均衡。以算例详细说明算法实现思路,在现实科学工作流条件下进行仿真测试,验证算法性能。与几种单一元启发式调度方法相比,验证该算法拥有更高执行效率和负载均衡度。Aiming at the cloud workflow scheduling problem,a workflow scheduling load balancing algorithm fusing genetic algorithm and particle swarm optimization algorithm was proposed.The advantage of multiple heuristic method fusion was utilized to avoid the slow convergence of genetic algorithm and that particle swarm optimization algorithm is easy to fall into local optimum.Workflow task was mapped effectively to the virtual machine resources,minimizing global workflow execution span and balancing virtual machine allocation of load.An example was used to specify the implementing idea of the algorithm.Under the reality of scientific workflow simulation test,the performance of the algorithm was verified.Compared with several single meta-heuristic workflow scheduling methods,the proposed algorithm has higher execution efficiency and load balancing degree.

关 键 词：虚拟机资源 工作流调度 云计算 元启发式方法 负载均衡 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]