网格计算中任务调度算法的仿真研究

摘要

The genetic algorithm in task scheduling has the defects of slow convergence speed and local mini-mum, resulting in poor network scheduling performance. A task scheduling model is proposed based on particle swarm algorithm to improve the scheduling efficiency. According to the principle and characteristics of particle swarm optimization, the task scheduling model establishes the grid scheduling of meta task model of performance index and mathematical model, then using particle swarm optimization algorithm to solve the model, resource utilization and the task execution efficiency. Simulation results show that based on particle swarm algorithm, the task scheduling strategy is feasible and efficient, improves the task scheduling speed and efficiency, and avoids the genetic algorithm of easily trapped into local optimal problem.%研究网格计算中任务调度优化问题,由于网格环境具有动态性、异构性等特点,对高效调试资源效率有影响,导致传统网格任务调度算法收敛速度慢、局部最优等缺陷,使网格任务调度效率低.为了提高网格任务调度效率,提出一种基于粒子群算法的任务调度模型.模型根据任务调度原理和粒子群算法特点,建立了网格任务调度的元任务模型和性能指标的数学模型,然后采用粒子群算法对该模型进行求解,提高资源利用率和任务执行效率.仿真结果表明,根据粒子群算法的任务调度策略,提高了任务调度的速度和效率,很好的解决网格任务调度中存在的难题.