中国电机工程学报2016，Vol.36Issue(19)：5111-5121,11.DOI:10.13334/j.0258-8013.pcsee.151863
一种电力设备状态监测大数据的分布式联机分析处理方法
A Method of Distributed On-line Analytical Processing of Status Monitoring Big Data of Electric Power Equipment
摘要
Abstract
ABSTRACT:The big data analysis of power equipment condition monitoring is a hot research topic, which has great significance to guarantee the safe and stable operation of electric power equipment. The on-line analytical processing (OLAP) is one of most important methods to realize big data analysis of status monitoring of electric power equipment. OLAP includes multi angle, quick access and analysis data. Through the analysis of OLAP technology in big data environment, a kind of analytical platform of status monitoring big data of electric power equipment was designed. This platform includes relational on-line analysis base on Hive, relational on-line analysis base on Impala, and multi dimensional on-line analysis based on HBase. Although distributed relational analysis has met the needs of power enterprises to some extent, the frequent connection between data model where the operation of data sheet, resulting in its performance bottlenecks. Aiming to solve the problems of large cost of connection operation and low query speed of distributed relational analysis data model, the report presented a kind of data schema of state monitoring of power equipment which was based on not-join level-encoding technologies. In order to reduce the number of connection options to optimize performance, encoded the level information of dimension table, compressed to the fact table. The big data analysis platforms of Hive and Impala were used in the data model, which include monitoring data in different scale load data of this model, roll-up and cost of storage for test. The experimental results show that compare with the conventional data model, in data loading respect, the loading rate of this model is increased 42%; in roll-up respect, the time is shortened by 40% to 49%; in storage cost respect, this model is about 37% less. This report verified the validity and feasibility of this model to bring the big data analysis of status monitoring of electric power equipment to a new train of thought.关键词
状态监测/大数据/关系型联机分析/数据模型/层次编码Key words
status monitoring/big data/relational on-line analysis/data model/level encoding分类
信息技术与安全科学引用本文复制引用
王德文,周青..一种电力设备状态监测大数据的分布式联机分析处理方法[J].中国电机工程学报,2016,36(19):5111-5121,11.基金项目
国家自然科学基金资助项目(61074078)。 Project Supported by National Natural Science Foundation of China (61074078)． （61074078）
