基于云模型的网格任务调度遗传算法研究  

Grid Task Scheduling Genetic Algorithm Based on Cloud Model

作　　者：郑世明[1] 高志年[1] 韦伟[1] 苗壮[2] 邵荣明 

机构地区：[1]南京陆军指挥学院作战实验中心,南京210045 [2]解放军理工大学指挥信息系统学院,南京210007 [3]解放军沈阳炮兵学院训练部,沈阳110867

出　　处：《电子科技大学学报》2012年第6期911-915,共5页Journal of University of Electronic Science and Technology of China

基　　金：国家863项目(2007AA01Z126);部级预研基金;江苏省自然科学基金(BK2012512)

摘　　要：针对网格环境动态多变性的特点,为了克服传统遗传算法易陷入局部最优的缺陷,提出了一种基于云模型的网格任务调度遗传算法。该算法由正态云模型的Y条件云发生器实现交叉操作,由基本云发生器实现变异操作,对调度模型进行优化求解,并在任务调度中对初始种群的产生、选择、变异和交叉操作进行了改进,通过实验分析,表明了该算法的可靠性、有效性和实用性。Aiming at the dynamic characteristic of grid, to overcome the shortcomings of genetic algorithms which easily get a local optimum solution, a cloud-based genetic algorithm （CGA） for grid task scheduling is proposed. CGA is based on both the idea of GA and the properties of randomness and stable tendency of a normal cloud model. In this algorithm, Y-conditional cloud generator is used for the crossover operator, and basic cloud generator is used for the mutation operator. CGA can optimizes the solution with genetic algorithm based on cloud-model, ascertains the oriental scenario for scheduling, and improves on the arithmetic operators of population initialization, select, crossover, mutation and reinsertion in the process of task scheduling. The experiment validates the feasibility, validity, and practicality of the algorithm.

关 键 词：云模型 遗传算法 网格 任务调度 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]