Optimizing day-ahead energy management with demand response in a PV-diesel-battery system using a hybrid GOA-SNN strategy

Abstract

This manuscript presents a hybrid technique for the day-ahead energy management strategy of hybrid renewable energy systems (HRES) integrated with demand response (DR). The HRES consists of Photovoltaic (PV), Diesel Generator (DG), and battery systems. The proposed method, named GOA-SNN strategy, combines the Gazelle Optimization algorithm (GOA) and spiking neural network (SNN) to improve outcomes for consumers and utility providers. The objective is to minimize human intervention and leverage advanced bi-directional communication. The focus of this study is on off-grid systems. The proposed strategy determines optimal incentives and power commands to maximize the operation of the HRES within a 24-hour control horizon. The GOA is used to optimize the scheduling of energy resources, considering constraints such as energy demand, battery state of charge, and grid availability. The SNN is used to predict energy demand using historical data and external factors. The strategy utilizes lower electricity prices to influence consumer behavior and identifies the ideal time periods to offer these discounted prices, optimizing the expected benefits of energy management. The proposed technique's performance is evaluated through MATLAB or Simulink simulations and compared against existing methods. The results demonstrate that the GOA-SNN strategy achieves cost savings and reduces computation time, making it an efficient and effective solution for energy management in hybrid renewable energy systems. The existing approach like BSSA, NPO, OSA and proposed technique accuracy becomes 70.23%, 75.59%, 89.42% and 95.21%. From this analysis, it concludes that the proposed method based accuracy is high compared with existing techniques.