HBase-based storage system for electrical consumption forecasting in a Moroccan engineering school

Abstract

Nowadays, several sectors expose expensive electrical consumption cost due to their high electrical needs. Educational institutions are among these buildings, due to their new practices and activities such as the use of electrical equipment, the implementation of complex scientific experiments, and the organization of big events in various domains. So, reliable energy forecasting system is required to manage future electrical budgets. The National School of Applied Sciences of El Jadida — Morocco has decided to change its energy policy, by installing a private smart grid based on photovoltaic panels that will cover 40% of its electricity needs, encourage local production and increase the share of renewable energies. According to the high level of complexity that smart grid data management presents due to the growth of data volumes coupled with the variety of data types and formats, the integration of Big Data technology is also required to resolve immoderate electricity consumption's forecast. In this paper, we suggest a Big Data based solution in term of data selection, integration and storage. In the first place, we select all factors that might have impact on electrical consumption (temperature, solar radiation, relative humidity, wind speed, equipment electrical features and occupancy schedule), then we propose a storage data model using HBase.