This paper presents a robust optimization model for unit commitment (UC) in smart power systems, aiming to enhance stability and reduce operational costs. The proposed model preserves the problem structure and security constraints while integrating renewable energy sources (RES), energy storage systems (ESS), and electric vehicles (EVs) within a robust optimization framework. Instead of traditional two-stage methods, an advanced genetic algorithm (GA) is employed, incorporating selection processes based on roulette wheel and tournament selection, uniform crossover, and variable mutation. This algorithm is specifically designed to identify the most optimal and cost-effective UC patterns under uncertainties arising from renewable generation fluctuations and load demand variations. Simulation results on the 6‑bus, 24‑bus, 118‑bus, and 300‑bus test systems implemented in Python demonstrate that the proposed approach significantly reduces operational costs, accelerates algorithm convergence, and enhances system resilience under critical conditions.