基于云计算任务调度的遗传粒子群优化算法

Genetic and Particle Swarm Optimization Algorithm Based on Cloud Task Scheduling

摘要: 云平台的任务调度算法是云计算领域研究的热点。如何在满足不陷入局部最优解的同时有更快的收敛速度，一直是研究者追求的目标之一。为此，本文提出将改进随机因子和惯性权重的增强型粒子群算法(EPSO)引入到遗传算法(GA)变异操作中的增强型遗传粒子群混合算法(GA_EPSO)。通过EPSO算法中的当前最优解和全局最优解重构变异算子，探索提升GA_EPSO算法在不陷入局部最优解的条件下，有更快的收敛速度。仿真实验表明，相同条件下，与遗传算法(GA)，改进后的遗传算法(IGA)，粒子群算法(PSO),增强型粒子群算法(EPSO)以及遗传粒子群算法(GA_PSO)相比，本文提出的改进算法不仅收敛速度快，而且任务调度效率也有明显提升。

Abstract: The task scheduling algorithm of cloud platform is a hot topic in the field of cloud computing. How to achieve faster convergence speed while not meeting the local optimal solution has always been one of the goals pursued by researchers. To this end, this paper proposes an enhanced genetic and particle swarm optimization algorithm (GA_EPSO) that introduces an enhanced particle swarm optimization algorithm (EPSO) with improved random factors and inertia weights into mutation operations in genetic algorithm (GA). Reconstructing the mutation operator by the current optimal solution and the global optimal solution in enhanced particle swarm optimization algorithm, the enhanced genetic and particle swarm optimization algorithm has a faster convergence speed without falling into the local optimal solution. Simulation experiments show that under the same conditions, compared with genetic algorithm (GA), improved genetic algorithm (IGA), particle swarm optimization (PSO), enhanced particle swarm optimization (EPSO) and genetic particle swarm optimization (GA_PSO), the algorithm not only accelerates the convergence speed, but also has a significant improvement in task scheduling efficiency.