Battery integration and grid-connected hybrid power plant with an optimal energy management system integrated into a multilevel configuration

A hybrid renewable energy system (HRES) is developed due to the variability of individual renewable energy sources (RES), which highly impacts their reliability. These systems are effective in meeting the demands of residential loads. However, an energy management system must address problems such as power surpluses and shortages. This manuscript presents an innovative hybrid method for ideal energy management in a grid-connected hybrid power plant integrated with a battery. The proposed hybrid method combines the use of the Snow Ablation Optimiser (SAO) and the Matrix Diffractive Deep Neural Network (MDDNN). It is therefore called the SAO-MDDNN method. The major goal of the proposed study is to extend the battery's lifespan. The Battery Energy Storage (BES) is optimized for efficiency using the SAO strategy. The MDDNN is utilized to predict the power demand. The proposed technique is put into practice in MATLAB and compared with traditional techniques and advanced methods like Position-Mutation Grey Wolf Optimization (PM-GWO), Mixed-Integer Linear Programming (MILP), and Long Short-Term Memory (LSTM). The proposed technique accomplished a high mean BES efficiency of 94.70 % and a low computation time of 6.47 s compared to the existing techniques. It also boasts a high cost reduction of 10.60 % and low CO2 emission of 149 kg/day. The proposed SAO-MDDNN method demonstrates significant potential for optimizing energy management in hybrid renewable energy systems, offering improved efficiency, cost reduction and environmental benefits.