电力大数据全景实时分析关键技术  

Key technology of power big data for global real-time analysis

作　　者：周国亮 吕凛杰 王桂兰[2] 

机构地区：[1]国网冀北电力有限公司技能培训中心,河北保定071051 [2]华北电力大学信息与网络管理中心,河北保定071003

出　　处：《电信科学》2016年第4期159-168,共10页Telecommunications Science

基　　金：河北省自然科学基金资助项目(No.F2014502069);中央高校基本科研业务费专项资金资助项目(No.13MS103)~~

摘　　要：针对智能电网建设过程中收集的电力大数据,基于电力系统全景实时数据分析的需求,探讨基于大数据的电力系统安全可靠性分析、实时状态监控及能源全景动态平衡调度等核心问题的解决思路。分析了利用大数据解决安全可靠性、设备全寿命周期管理及能源实时平衡调度等问题的挑战及解决思路,基于大规模实时多源细节数据和设备全景数据的计算,有助于提高系统分析的精度和准确度,保证电网安全运行;探讨了内存计算、实时流式大数据处理、大规模并行计算及列存储等技术在电力大数据实时分析中的应用;结合主流开源大数据处理技术,设计了电力大数据分析平台的分层体系架构,为电力系统的高效运行提供保证。For power big data collected during smart grid construction process, based on the demand of power system global and real-time data analysis, ideas of solving power system security and reliability, real-time status monitoring, energy global dynamic balance scheduling and other key issues were explored. The problems of big data safety and reliability, equipment life-cycle management and energy real-time balance scheduling were analyzed and discussed, system analysis precision and accuracy based on large-scale real-time multi-source detail data and global data of equipment would be improved, then application of in-memory computing, real-time streaming data processing technology, massively parallel computing technology and column stores were explored; a layered architecture of power big data analytics platform which combined with the mainstream open source big data processing technology was proposed to provide guarantees for the efficient operation of the power system.

关 键 词：电力大数据 全景实时数据 内存计算 数据流 大规模并行 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]