Deep learning based optimal energy management framework for community energy storage system

Abstract

This paper proposes a deep learning-based integrated framework for multiple cooperative households to achieve optimal energy distribution. The corresponding energy generation and consumption problems are formulated by a long short-term memory algorithm is combined with an optimization algorithm to produce an optimal solution. In this study, a PV-community energy storage system (CESS) integrated is considered where the scheduling decision of the CESS and utility grid can be subsequently achieved through formulated constraints. The test results demonstrate the efficacy and robustness of the proposed system that achieves superior performance on effective renewable energy usages of maximum 31.74% in a home environment.

Keywords

• Energy storage system
• Long short term memory
• Optimization algorithm
• Home energy management system