云平台下全维度电力设备监测数据并行化处理技术

Parallel processing technology of full dimension power equipment monitoring data under cloud platform

数据集成和信息共享是建立智能电网的必然趋势,设备状态数据越来越多地被发送到控制中心,如何快速处理大量的历史数据和实时在线监测数据成为亟待解决的问题。结合开源云平台和大型数据处理技术,对电力设备运行状态监控数据进行并行计算和诊断研究,提出基于Spark内存技术的集合经验模型的分解并行算法EEMD,以补偿复杂场景处理中HDoop Map Reduce的不足,设计实现了两种不同结构的并行EEMD算法,并通过对比实验分析算法的性能,研究工作可为我国电力设备状态数据并行处理技术的发展提供一定的参考和借鉴。

Data integration and information sharing are the inevitable trend of building smart grid.More and more equipment status data are sent to the control center.How to quickly process a large number of historical data and real-time online monitoring data has become an urgent problem.Combining with open source cloud platform and large data processing technology,the parallel computing and diagnosis of power equipment running state monitoring data are studied in this paper.A decomposition parallel algorithm EEMD based on Spark memory technology is proposed to compensate for the deficiency of HDoop Map Reduce in complex scene processing.Two different structures are designed and implemented.The parallel EEMD algorithm and the performance of the algorithm are compared and analyzed.The research work can provide some reference for the development of the parallel processing technology of the state data of power equipment in China.