Energy forecasting using multiheaded convolutional neural networks in efficient renewable energy resources equipped with energy storage system

Abstract

Renewable energy resources (RERs) motivate electricity users to reduce their energy bills by taking benefit of self-generated green energy. Different studies have already pointed out that, because of the absence of proper technical support and awareness, the energy users were not able to sufficiently take paybacks from the RERs. However, with the commencement of smart grids, the potential benefits of RERs and dynamic pricing schemes can be exploited. Nonetheless, the big issue is the accurate prediction of energy produced by intermittent RERs. In this work, we have proposed an efficient framework by integrating energy storage system (ESS) and RERs with smart homes. This framework has shown significant results, which make it helpful and suitable for energy management at a community level. We applied a multiheaded convolutional neural network model for precise and accurate prediction of produced energy by RERs. Moreover, we have considered a smart community consisting of 80 homes. Simulation results prove that the proposed framework helps to decrease the energy bill of consumers by 58.32% and 63.02% through integration of RERs without and with ESS, respectively.