摘要
针对云计算中现有智能任务调度算法容易陷入局部最优的问题,提出一种基于改进型离散粒子群优化(DPSO)算法的任务调度方案。对传统DPSO算法中的粒子位置更新公式中的惯性权重进行改进,使其根据迭代次数非线性递减,提高算法的搜索能力;另外,融入了随机扰动操作,避免算法陷入局部最优。实验结果表明,与传统遗传算法和粒子群算法相比,该方案能够获得最优的调度策略,有效降低任务的完成时间。
For the issues that the existing intelligent task scheduling algorithm is easy to fall into local optimization in cloud computing, a task scheduling scheme base on the improved discrete particle swarm optimization（DPSO） algorithm is proposed.The inertia weight of the particle position update formula in the traditional DPSO algorithm is improved, so that nonlinear decreasing according to the number of iterations,to improve search ability of the algorithm. In addition, the stochastic perturbation operation is integrated into the update formula to avoid algorithm trapped in local optimum. Experimental results show that compared with the traditional genetic algorithm and particle swarm optimization, the proposed scheme can obtain the optimal scheduling strategy, which can effectively reduce the completion time of the task.
出处
《齐齐哈尔大学学报（自然科学版）》
2017年第4期6-10,共5页
Journal of Qiqihar University(Natural Science Edition)
关键词
云计算
任务调度
离散粒子群优化
惯性权重
随机扰动
cloud computing
task scheduling
discrete particle swarm optimization
inertia weight
stochastic disturbance
