Abstract: This paper analyzes the task features of ensemble prediction, based on the Web Services Resource Framework(WSRF), it designs a dynamically adaptive ensemble prediction task schedule system. The adaptive schedule algorithm is proposed, according to the dynamic changes of grid nodes, it assigns the optimal number of CPU for model prediction task dealing with initial samples. The schedule system can save computational costs, while shorten the time of prediction. The practice operation results demonstrate the feasibility of the system.

Key words: Web Services Resource Framework(WSRF), task schedule, ensemble prediction, model prediction

摘要： 分析集合预报的任务特点，基于Web服务资源框架(WSRF)设计一种动态自适应的集合预报任务调度系统。给出任务调度算法，根据网格节点的资源变化自适应地选择节点，为每对初值样本的模式预报任务指派CPU资源数目，以缩短预报时间、节省计算成本。实际运行结果验证了该系统的可行性。

关键词: Web服务资源框架, 任务调度, 集合预报, 模式预报

CLC Number: 

• TP391