Electricity Demand Forecasting Models at Hourly and Daily Level: A Comparative Study

Abstract

Due to industrialization and an increase in population, the electricity demand has increased sharply. There is a gap between the supply and requirement of electricity. Electricity forecasting plays a very significant role in power grid as it is required to maintain balance between supply and load demand at all the times, to provide a quality supply of electricity, for financial planning, generation reserve, system security, and many more. Forecasting power is one of the complex problems due to various factors like time and weather. It becomes easier to store relevant data due to technological advancements (Smart Home and Internet of Things-IoT). The electricity consumption data collected through sensor devices can be utilized to know future electricity requirements. In this paper we have applied ten models on the Electricity consumption dataset of house from 11 Jan, 2016, to 27 May 2016 (around 4.5 Months duration) per 10-minute observation. It is observed from the results that Facebook Prophet model is the best performing model.