Supply and Value Estimation of Electricity in Smart Grids using Convolutional Neural Networks and Big Data Analytics

Abstract

Electrical power systems are evolving into more intelligent and green frameworks. These tendencies have several
beneficial outcomes, including the active engagement of consumers in power marketplaces. Nevertheless, the resultant
consumption-side flexibilities induce significant demand variations, complicating the maintenance of power
equilibrium and dependability inside smart grids (SG). Users may modify their power consumption behaviors in SG
due to electricity pricing and subsidies. This leads to prices that may deviate from the original pricing structure.
Predicting electricity supply and value is essential for SG's dependability and ecological viability. The use of big data
analytics (BDA) for predicting has emerged as a prominent study area due to the substantial volume of data created
and stored in the SG setting. With enhanced awareness of pricing and demand, power consumers may optimize their
load management effectively. This research employs a Convolutional Neural Network (CNN) and BDA for SG Supply
and Value Estimation of Electricity (SVE). Consequently, the proposed model uses many variables to forecast
prospective energy usage and value metrics. The configuration of variables is essential since the efficacy of the
prediction model is contingent upon these values.