Application of particle swarm optimization algorithm based on classification strategies to grid task scheduling

Grid task scheduling is a NP-hard problem. Inthis paper, an optimization algorithm of grid taskscheduling is brought forward by using classificationstrategies to improve particle swarm algorithm. The particleswarm is divided into accurate subgroups for local slowsearch, commonness subgroups for the cloning strategyprocessing and inferior subgroups for changing intoaccurate subgroups to operate the positive and reverseclouds. The experimental results show that the schedulingalgorithm effectively achieves the load balancing ofresources and preferably avoids falling into local optimalsolution and the selection pressure of genetic algorithm andelementary particle swarm algorithm. This algorithm hasthe high accuracy and convergence speed and so on. © 2012 ACADEMY PUBLISHER.