云计算环境下基于改进粒子群算法的任务调度

Task scheduling based on improved particle swarm optimization in cloud computing environment

为了优化云计算环境下任务调度,考虑调度过程中任务的最短完成时间、系统的负载均衡和经济成本3个目标约束,然而3个目标约束之间存在冲突,因此提出了一种使用改进粒子群优化算法来解决云计算任务调度中多目标优化问题,达到同时兼顾3个目标约束的目的。选择惯性权重的模糊自适应策略对粒子群算法进行改进,从而能很好的平衡粒子的全局搜索能力和局部搜索能力,尽量避免过早收敛和陷入局部极值,并且引入移动子和负载因子的概念,用于实现算法对云计算环境下的任务调度。仿真结果表明,该算法对多目标优化问题,具有较好的寻优能力。

For optimizing the task scheduling of cloud computing environment,it processes to consider the shortest completion time, load balancing, system constraints and economic costs of the three objectives, however, there is still have a conflict between these three objectives of constraints, thus we propose a method of using improved Particle swarm optimization（PSO）algorithms to solve the purpose of cloud computing task scheduling for multi-objective optimization problem to consider the three objectives constraints. Improving the global search ability and local search capability by using Fuzzy Adaptive Inertia Weight PSO strategy so that the particles can be well balanced to avoid premature convergence and local extremum,introducing the concept of the moving element and the load factor for the realization of task scheduling algorithm for cloud computing environments.Simulation results show that the algorithm for multi-objective optimization problem has better search capability.