面向云平台的免疫多目标优化调度算法  

IMMUNE MULTI-OBJECTIVE OPTIMIZATION SCHEDULING ALGORITHMFOR CLOUD PLATFORM

作　　者：李颖 邵清[1] 王清雲 周子航 夏凤阳 Li Ying;Shao Qing;Wang Qingyun;Zhou Zihang;Xia Fengyang(School of Optical Eleclrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《计算机应用与软件》2024年第1期278-284,共7页Computer Applications and Software

基　　金：国家重点研发计划项目(2018YFB1700902)。

摘　　要：云环境下任务之间存在多种特征,由于传统的资源分配机制存在变化和不确定等特征,容易引发负载不均衡使得调度受到制约,任务时延约束也会降低任务调度策略的利用率。针对这些问题,提出一种面向云平台的免疫多目标优化调度算法。利用Pareto支配关系,设计出云计算任务调度问题的数学模型;经过种群初始化、获得Pareto最优解、计算拥挤距离、克隆选择、重组和变异一系列操作,保持种群的多样性,实现调度的全局优化。与传统算法进行对比,实验结果表明该算法的搜索范围更广,在解的搜索广度上更加优秀,并且还有效平衡了任务执行时间和执行费用,提高了用户满意度。In the cloud environment,there are various characteristics among tasks.Due to the changes and uncertainties in the traditional resource allocation mechanism,load imbalance is easy to cause scheduling constraints,and task delay constraints also reduce the utilization of task scheduling policies.To solve this problem,a cloud-oriented platform immune multi-objective optimization scheduling algorithm is proposed.Pareto dominance relation was used to design the mathematical model of cloud computing task scheduling problem.After population initialization,Pareto optimal solution,calculation of crowding distance,clone selection,recombination and variation,the diversity of population was maintained and the global optimization of scheduling was realized.Compared with the traditional algorithm,the experiments show that the proposed algorithm has a wider search range,better search breadth of solutions,and can balance the task execution time and cost effectively,and improve user satisfaction.

关 键 词：云平台 人工免疫 多目标 任务调度算法 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]