针对当前网格工作流调度算法中大多只考虑DAG结构的网格工作流、涉及QoS参数较少及将多QoS参数聚合成一个单目标函数进行优化调度的现状,提出了一种新颖的网格工作流调度算法。该算法基于表达结构丰富的AGWL语言建模网格工作流,且基于MOPSO算法所设计的带多QoS约束的多目标优化的网格工作流调度算法。通过与基于NSGA-Ⅱ算法的网格工作流调度算法比较,表明了该算法的有效性。
Existing grid workflow scheduling algorithms commonly are suffered by one or several of the following drawbacks: Most only have considered the structure of the DAG grid workflow,have not enough considered the multidimentional QoS parameters and the multidimentional QoS parameters will be aggregated into a single objective function for optimal scheduling.The paper presents a novel grid workflow scheduling algorithm.The algorithm is based on the structure of the expression of a rich language AGWL grid workflow modeling, and is based on MOPSO algorithm designed with multiple QoS constraints of muhiobjective optimization of the grid workflow scheduling algorithm.The proposed algorithm is compared with a grid workflow scheduling algorithm based on the NSGA-Ⅱ algorithm, the experiment results show the validity of algorithm.