云计算环境下基于改进离散粒子群的并行调度算法  

Improved Discrete Particle Swarm-Based Parallel Schedule Algorithm in Cloud Computing Environment

作　　者：徐华[1] 张庭[1] 

机构地区：[1]江南大学物联网工程学院,江苏无锡214122

出　　处：《华南理工大学学报（自然科学版）》2015年第9期95-99,共5页Journal of South China University of Technology(Natural Science Edition)

基　　金：国家留学基金委资助项目(201308320030);江苏省自然科学基金资助项目(BK20140165)~~

摘　　要：针对云计算环境下的任务调度优化问题和传统离散粒子群优化(DPSO)算法早熟、精度低等缺点,提出了一种适合云计算环境下动态调整惯性权重因子的方法,并给出了云计算环境下改进后的离散粒子群优化算法.该算法能快速确定合适的并行任务分配方案,使其达到调度长度最短的优化目标.仿真结果表明:文中改进的DPSO算法的收敛性、前期全局搜索和后期局部探索性能均优于传统的DPSO算法和遗传算法;在任务数较大的情况下,采用改进DPSO算法的并行任务调度算法的调度长度明显优于采用传统DPSO算法和遗传算法的并行任务调度算法.Aiming at the optimization problem of task scheduling in the cloud computing environment and the de-fects of prematurity and low precision of traditional discrete particle swarm optimization （DPSO）algorithms,a method of dynamically adjusting the inertia weight factor is proposed in a cloud computing environment,and an im-proved discrete particle swarm optimization algorithm is put forward.This algorithm can determine the appropriate parallel task allocation scheme quickly,and makes the scheme achieve the shortest scheduling length.Simulation results show that the improved DPSO algorithm is superior to the traditional DPSO algorithm and the genetic algo-rithm in terms of the convergence,the previous global search capability and the late local exploration performance, and that,in the case of a large number of tasks,the parallel task scheduling algorithm using the improved DPSO algorithm is superior to those using the traditional DPSO algorithm or the genetic algorithm in terms of scheduling length.

关 键 词：云计算 并行算法 离散粒子群优化 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]