| As the important foundation of power quality monitoring hierarchy, the county-level power quality monitoring system has numerous monitoring nodes and a large amount of data. With the increase of monitoring nodes and data, a trend of massive data has been showed, which brings big pressure to the county-level power quality monitoring system and the the cost of system expansion and labor will rise sharply. As the limitation of conditions, the county-level power supply enterprises can only invest limited resources,therefore, it's very meaningful for them to make full use of existing resources to build a power quality monitoring system with efficient storage and fast processing.The problems of data parsing, storage and query in power quality monitoring system are studied in this paper. Firstly, by studying the worldwide development situation of power quality monitoring system and NoSQL, the advantages of MongoDB in dealing with the massive data are pointed out. Through the analysis of county-level power quality monitoring system's demand, two key problems in the system has been cleared, and a massive power quality data storage solution based on MongoDB is put forward and implemented with a set of power quality data interface. Then, the insufficiency of the existing power quality data paring solution is analyzed, combining the Hadoop MapReduce parallel framework, a fast data parsing solution for county-level power quality monitoring system is proposed and implemented. Finally, in order to verify the performance of storage solution, parsing solution and comprehensive solution, writing and query performance of storage solution, data parsing performance of parsing solution and the performance of comprehensive solution are tested and analyzed.The test results show that the proposed solution has better data storage and query performance compared with the solution based on relational database, and a higher efficiency compared with the the existing data parsing solutions. |
PDF Full Text Request