Abstract
wind turbine and battery energy storage units for supplying the green electricity to the electric vehicle (EV)
charging infrastructure in Delhi. It discusses the critical issues which involves the detrimental effect on the power
grid due to the rising demand of EVs, fluctuating behavior of the RES, insufficient integration of energy storage
devices and the inadequacies of the classical optimization methodologies in handling intricate and multi-faceted
system development. A thorough technological and economic investigation is carried out to examine various
system configurations. The main goal is to find the best sizes for decision variables to lower energy costs, TNPC,
and the chance of power outages. To reach these goals, the Artificial Hummingbird Algorithm (AHA), which is a
type of metaheuristic optimization method is employed. The AHA’s performance is contrasted with other optimization
methods, such as the Bald Eagle Search Algorithm (BESA), the Salp Swarm Algorithm (SSA), and the
Cuckoo Search Algorithm (CSA), by running 50 simulations with different levels of power supply deficit probability
(DPSP) of 0%, 1%, 3%, and 5%. It is observed that, AHA is more exact and consistent than BESA, SSA, and
CSA, with TNPC reductions of 7.9%, 14.1%, and 17.9%, respectively. Also, AHA reduces the LCOE to $0.3697/
kWh, which is 35.3%, 45.7%, and 57.8% less than the LCOE of BESA, SSA, and CSA, respectively. The financial
study exhibited that the SPV/WT/BESS-based EV charging system has the lowest payback period, which is 6.67
years. It also has a 6.8% annualized return on investment (AROI), which is better than the other system configurations.
The AHA shows that combining solar, wind, and battery storage systems makes EV charging infrastructure
much more reliable, cost-effective, and ecologically approachable. This study is novel because it utilizes
AHA for the first time to improve multifaceted, multi-objective renewable energy-based EV charging systems.
Further, the proposed technique leads to fast converging behavior and solution quality than other existing
techniques. The main contribution of this study is the application of an AI-driven, location-specific optimization
paradigm that assist the large-scale rollout of clean and efficient EV charging solutions.