邓松,岳东,朱力鹏,胡斌,周爱华.电力大数据智能化高效分析挖掘技术框架[J].电子测量与仪器学报,2016,30(11):1679-1686
电力大数据智能化高效分析挖掘技术框架
Framework of intelligent and efficient analysis and mining technology for power big data
  
DOI:10.13382/j.jemi.2016.11.008
中文关键词:  电力大数据  数据挖掘  特征提取  多源数据过滤
英文关键词:power big data  data mining  feature extraction  multi source data filtering
基金项目:国家自然科学基金(51507084)、中国博士后基金(2016M591890)、南京邮电大学引进人才基金(NY214203)、国家电网公司科技资助项目DOI: 10.13382/j.jemi.2016.11.008
作者单位
邓松 南京邮电大学先进技术研究院南京210023 
岳东 南京邮电大学先进技术研究院南京210023 
朱力鹏 全球能源互联网研究院计算及应用研究所北京102209 
胡斌 全球能源互联网研究院计算及应用研究所北京102209 
周爱华 全球能源互联网研究院计算及应用研究所北京102209 
AuthorInstitution
Deng Song Nanjing University Post & Telecommunication, Nanjing 210023, China 
Yue Dong Nanjing University Post & Telecommunication, Nanjing 210023, China 
Zhu Lipeng State Grid Smart Grid Research Institute, Beijing 102209, China 
Hu Bin State Grid Smart Grid Research Institute, Beijing 102209, China 
Zhou Aihua State Grid Smart Grid Research Institute, Beijing 102209, China 
摘要点击次数: 122
全文下载次数: 98
中文摘要:
      随着智能电网的不断建设,各类生产经营管理活动中产生的海量、高频数据,具有实时性、易失性、突发性、无序性、无限性等特征,如何充分利用和分析这些数据,快速获取有价值的信息是当前电力大数据处理急需突破的难点。在分析目前电力大数据应用现状的基础上,构建了电力大数据智能化高效分析挖掘技术框架,同时从面向计算密集型电力大数据的特征分析技术、基于内存计算的高性能数据分析技术、电力大数据并行化分析框架及服务体系以及基于数据挖掘的母线超短期负荷预测技术4个方面详细描述了电力大数据智能化高效分析挖掘的关键技术,从而为电力业务数据的高效价值挖掘及在线决策分析提供理论依据及基础技术支撑。
英文摘要:
      With the development of smart grid, mass and high frequency data was produced, which has the characteristics of real time, volatile, burst, disorder, infinite and so on. How to use and analyze these data, access to valuable information is an urgent problem to be solved in the analysis of power big data. On the basis of analyzing the current situation of the application of power big data, this paper constructs the framework of intelligent and efficient analysis and mining technology for power big data. Meanwhile, the key technologies of intelligent and efficient analysis and mining technology for power big data are described in detail from feature analysis technique for computing intensive power big data, high performance data analysis technology based on memory computing, framework and service system for parallel analysis of power big data and super short term load forecasting based on data mining. However, theoretical basis and technical support is provided for the efficient value mining and online decision analysis of power business data.
查看全文  查看/发表评论  下载PDF阅读器