摘要
提出了一种云计算环境下的电力数据在线安全分析并行化新方法,提出了该方法的电力数据分析的数学模型、子任务划分的方式;算法采用MapReduce编程平台,利用Hadoop的HDFS(Hadoop Distributed File System)来存储大容量的电网数据;描述了MapReduce的电力数据并行处理的工作机制与流程;通过Map和Reduce这种主-从编程模式很方便使电力在线安全分析的子任务在Hadoop的PC集群上运行。IEEE118节点的电网作为电力数据在线安全分析的电网数据,测试结果表明:针对大规模电力系统在线安全分析快速计算需求,该方法具有较好的执行时间与加速比。
A parallel process approach in power system on-line security analysis based on cloud computing was presented in this paper. The mathematic model and sub-task partition of this large scale power flow security analysis were also discussed subsequently. It used the Map/Reduce distributed computation programming model and thus the large power data was stored in HDFS （Hadoop Distributed File System）. The working mechanism and the flow of power system on-line security analysis on Map/Reduce was also discussed and described in detail. Using Map and Reduce manipulation, our on-line security analysis algorithm can easily run on Hadoop PC clusters by a master-worker style. IEEEll8 node power system is used for the performance testing of on-line security analysis application and the results demonstrate that our algorithm can meet the fast demands in large scale electric power systems. Good execution time and speedup are also obtained in this test.
出处
《控制工程》
CSCD
北大核心
2017年第4期823-828,共6页
Control Engineering of China
基金
国家自然科学基金(1300180)
中央高校基本科研业务费专项资金资助项目(TD2014-01)
