Statistically Validated EV Charging and Discharging Scheduling Optimization in V2G System: A Comparative Evaluation of Multiple Metaheuristic Algorithms

The rising uptake of electric vehicles (EVs) is a critical challenge for how power systems function, as it places a strain on the grid, especially during peak times, and as more EVs come into the system, uncoordinated charging demands will massively increase the stress on the grid. In this research, seven metaheuristic algorithms, including the Particle Swarm Optimization (PSO), the Differential Evolution (DE), the Whale Optimization Algorithm (WOA), the Grey Wolf Optimizer (GWO), the Enhanced Whale Optimization Algorithm (EWOA), the Particle Swarm Optimization–Whale Optimization Algorithm (PSO-WOA), and the Adaptive Particle Swarm Optimization (APSO), are used for the optimum scheduling of EV charging and discharging in the vehicle-to-grid (V2G) environment, and they are also tested statistically. The approaches are assessed using a realistic time-of-use (ToU) electricity pricing scheme, with peak, mid-peak, and off-peak time zones. The simulation results clearly demonstrate that APSO achieves the lowest mean value of the composite scheduling objective among the seven tested algorithms, improving on PSO by 8.0% and on the PSO-WOA hybrid by 9.7%, as well as showing a measurable improvement in the peak load demand profile. Moreover, PSO-WOA is not statistically different from PSO, and EWOA has not been shown to be better than WOA after family-wise correction (FWC) for seven algorithms. Thorough statistical validation, including parametric tests (t-test, ANOVA), non-parametric tests (Mann–Whitney U, Wilcoxon Signed-Rank, and Friedman) and post hoc analyses (Holm’s Step-Down, Bonferroni–Dunn, and Nemenyi), is done to distinguish true performance gains from random error. The results indicate that, for this problem class, the effective mechanism is adaptive control of swarm parameters rather than hybridization of search operators.