Research And Application On Technologies Of Big Data Integration And Management In Power Dispatching And Controlling

With the continuous expansion of the service scope of power dispatching and controlling system, an increase in the data source, data types, data size and processing pressure increases rapidly, massive and multi-source data analysis processing requirements for dispatching and controlling business are continuously advancing.This article, based on the big data of power dispatching and controlling system, is aiming at the characteristics of multi-source heterogeneous data of smart grid, and exploring the method of data integration and management in power dispatching and controlling system.This article proposes a unified data model of power system, which is the theoretical and practical basis for the data integration and management of power dispatching and controlling. To establish a unified data model of the power system, we can analyze the impact of big data for dispatching business, and define the power business requirement model under the environment of big data, electric power business demand model, and reflect the power dispatching and controlling business through data and relations, and finally achieve the multi-source heterogeneous data integration and management in grid business.In the data integration process, the main research is the ETL operation on the data, that is, data extraction, transformation and loading. Taking into account the data quality requirements, we will also study the technology of data cleansing and conversion.To study the technology of data storage and management, we can develop a suitable storage mode for different types of data. The distributed storage technology and relational database are both used to manage the data of power dispatching and controlling. This article also uses the zipper table to store the slowly varying data, which can store data in history version and save the storage space in the same time.Data integration and management in power dispatching and controlling, whic h is the basis and support for big data research, is a necessary part in research on power dispatching and controlling.