基于混合粒子群算法的云计算任务调度研究

任务调度是云计算系统可靠运行的关键,云计算环境中要处理的任务量巨大,考虑到云计算任务调度和QoS的优化问题,提出一种混合粒子群优化算法用于云任务调度。算法中引入遗传算法的交叉和变异思想,并结合随迭代次数变化的变异指数,保证种群进化初期具有较高的全局搜索能力,避免出现“早熟”,同时将爬山算法引入粒子群算法,改善局部搜索能力。实验结果显示该算法具有很好的寻优能力,是一种有效的云计算任务调度算法。

Cloud Task Scheduling Based on Hybrid Particle Swarm Optimization Algorithm

Task scheduling is the key to run cloud computing system reliably, huge task is to process in cloud computing environment, considering the optimization problem of cloud computing task scheduling and QoS, a hybrid particle swarm optimization algorithm for cloud task scheduling is proposed. The thought of crossover and mutation in genetic algorithm is introduced, and combined with the variance index changes with the number of iterations, to guarantee relatively high global search ability in initial stage of population evolution, and to avoid the "premature", at the same time, hill-climbing algorithm is introduced into particle swarm algorithm to improve the local search ability. The experimental results show that the algorithm has good optimization ability, and it is a kind of effective cloud computing task scheduling algorithm.