Reliable solar PV on-site generation for EV charging management in commercial buildings using LBO-DTRSRN approach

Effective energy management is crucial for commercial buildings equipped with solar photovoltaic (PV) panels and EV charging infrastructure, particularly due to the unpredictable departure timings of EV users. Traditional building energy management systems often fail to accommodate these variable behaviors, resulting in suboptimal performance and user dissatisfaction. To address this issue, this manuscript introduces a novel hybrid methodology for optimizing solar PV on-site generation and EV charging management in commercial settings. The proposed LBO-DTRSRN approach integrates the ladybug beetle optimization (LBO) algorithm with the double transformer residual super-resolution network (DTRSRN). This innovative method utilizes LBO to optimize charging schedules and energy usage, while DTRSRN accurately predicts EV charging demand. Performance evaluation of the LBO-DTRSRN approach was conducted using MATLAB and compared against methods such as long short-term memory (LSTM), particle swarm optimization (PSO) and bounded real-time dynamic programming (BRTDP) algorithm. Results demonstrate that the LBO-DTRSRN approach achieves a 17.8% reduction in operational costs, surpassing the reductions offered by PSO (14.7%), LSTM (16.7%) and BRTDP (15.8%). This study highlights the effectiveness of the LBO-DTRSRN approach in minimizing operational costs, providing a robust solution for optimizing energy management in commercial buildings with solar PV and EV charging infrastructure.