A Big Data–Driven Data Modeling Framework for Smart Energy Platforms

Abstract

The intensive development of smart energy systems has resulted in the continual production of high volumes of heterogeneous data of smart meters, sensors, grid infrastructure, and distributed energy resources. The management of this data must be effective to achieve real-time monitoring, predictive analytics and intelligent decision-making in the smart energy platforms. In this paper, a data modeling framework using big data is introduced to facilitate data mapping, integration, and management of high volume, velocity and variety energy data. The suggested framework follows a layered architectural design in order to process structured, semi-structured, and unstructured data and guarantee the scalability and flexibility of the system. The big data technologies are used to offer distributed storage, fault tolerance and parallel processing features, which increases efficiency of data access and lowers processing latency. The framework is tested with representative energy datasets and performance analysis is performed in terms of scalability, data retrieval performance as well as processing performance. According to the results of the experiment, the suggested data modeling framework proves to be much more efficient in the organization of the data and the provision of sophisticated analytics to serve smart energy exploitation applications (demand forecasting, energy monitoring, and operational optimization).