一种改进的粒子群算法在云计算资源中的研究  

Research of an improved particle swarm algorithm in Cloud computing resources

作　　者：史振华[1] 

机构地区：[1]绍兴职业技术学院,浙江绍兴312000

出　　处：《微型机与应用》2017年第6期62-65,共4页Microcomputer & Its Applications

摘　　要：针对云计算中无法合理分配资源的问题,提出了采用改进的粒子群算法对其进行分配。由于粒子群算法存在局部收敛速度快,容易陷入局部最优解的情况,从粒子的选择、参数调整和预防过早收敛三个方面进行改进:(1)选择粒子的时候采用适应值最小的粒子,并根据约束函数淘汰不合格的粒子;(2)对惯性因子、个体最优因子和全局最优因子进行改进;(3)通过设定系数来防止算法过早收敛。仿真说明,将改进后的算法运用在云计算方面,在云计算的资源失效、云端用户完成时间和云计算环境下的能量消耗三个方面都优于粒子群算法。Aiming at the issue that Cloud computing cannot reasonably allocate resources,this paper proposes adopting improved particle swarm algorithm to distribute resources. First of all,this paper makes the improvement from three aspects： the selection of particles,parameter adjustment and preventing premature convergence：（1） while choosing particles,it adopts particles with the minimum adaptive value,and eliminates disqualified particles according to constraint functions;（2） it makes the improvement of inertia factor,optimal individual factor and global optimal factors：（3） it prevents premature convergence of the algorithm through setting coefficient. The improved algorithm is applied to Cloud computing＇s resource scheduling,and achieves good effects in three aspects： failure of Cloud computing resources,completion time of Cloud-end users and energy consumption in Cloud computing environment.

关 键 词：粒子群算法 云计算 资源分配 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]