面向行政管理系统的任务调度算法研究

Research on task scheduling algorithm for system of administration

为了更好地解决云计算在行政管理系统任务调度交互数据量大的问题,本研究在综合遗传算法与退火算法的优点基础上提出了一种遗传退火混合算法。该算法针对行政管理系统中的任务调度数据关系特点,在云计算环境下对执行任务进行整数编码,通过遗传算法的适应度函数对任务数据进行全局最优解的搜索,任务数据的编码交叉重组过程中,利用退火算法调整适应函数的方式加快系统虚拟机对任务调度的完成时间。实验表明,该任务调度算法适应度值高于传统遗传算法且具有较强的搜索能力和收敛性。

To better address the cloud computing system administration task scheduling problem of large interactive data, this study based on the advantages of integrated genetic algorithms and annealing algorithm presents a genetic annealing hybrid algorithm. The algorithm for the administrative task scheduling system characteristic data relationships in a cloud computing environment to perform tasks integer coding genetic algorithm fitness function of task data to search the global optimal solution, coding task data crossover recombination process, function of the ways to speed up the system on using annealing algorithm to adjust the adaptive a virtual machine task scheduling is complete Experiments show that the task scheduling algorithm fitness value is higher than the traditional genetic algorithm and has a strong search ability and convergence.