A Data-Driven Approach Towards Forecasting Generalized Mid-Term Energy Requirement for Industrial Sector Users of Smart Grid

摘要

One of the major improvements that Smart Grid offers over traditional power grid is a balanced supply demand ratio. As electricity is hard to store for future usage, it is important to be aware of the demand in order to generate enough electicity for uninterrupted power supply. Thus, forecasting plays a vital role in Smart Grid. However, with various range of rapidly fluctuating parameters that influence electricity consumption patterns, it is next to impossible to design a single forecasting model for different types of users. Typically, electricity usage depends on demographic, socio-economic and climatic environment of any region. Besides, the dependencies between influencing parameters and consumption varies over different sectors, like, residential, commercial and industrial. In this paper, our main goal is to develop a generalized mid-term forecasting model for industrial sector, that can accurately predict quarterly energy usage of a large geographic region with diverse range of influencing parameters. The proposed model is designed and tested on real life datasets of industrial users of various states in the U.S.