A robust optimization framework for energy management in smart EV parking lots

The increasing penetration of electric vehicles (EVs), renewable energy sources, and uncertain electricity market conditions has created significant challenges for efficient energy management in smart EV parking lots. This paper proposes a hybrid multi-objective optimization framework, termed Hybrid Self-Improved Seagull Optimization Algorithm with Differential Evolution (H-SI-SOA-DE), for short-term energy management of EV parking lot systems under renewable generation, electricity price, and EV mobility uncertainties. The proposed framework simultaneously optimizes four conflicting objectives: total operational cost, voltage deviation, average EV fleet state-of-charge (SOC), and peak grid power dependency. A chance-constrained optimization formulation is developed to incorporate uncertainties associated with wind generation, electricity pricing, and EV arrival patterns while maintaining system operational feasibility. The proposed H-SI-SOA-DE framework integrates adaptive exploration mechanisms from Self-Improved Seagull Optimization Algorithm (SI-SOA) with Differential Evolution (DE)-based refinement to achieve an effective balance between global exploration and local exploitation. The algorithm is evaluated using the IEEE-33 bus distribution system and compared with benchmark optimization techniques, including NSGA-III, MOPSO, and classical SOA. The results obtained from 20 independent simulation trials demonstrate that the proposed framework achieves superior performance by reducing operational cost to ₹5.22 × 10 3 , minimizing voltage deviation to 0.051 p.u. improving average EV fleet SOC to 91.5%, and reducing peak grid power dependency to 1.88 MW. Compared with the non-optimized case, the proposed approach achieves a 17.7% reduction in operational cost, a 21.5% reduction in voltage deviation, an 11.5% improvement in EV fleet SOC, and a 36.3% reduction in grid dependency. Furthermore, statistical robustness analysis, sensitivity evaluation, scalability assessment with increasing EV penetration, and computational performance analysis demonstrate the reliability and adaptability of the proposed framework under uncertain operating conditions. The results indicate that the proposed H-SI-SOA-DE framework provides an effective and flexible solution for short-term smart EV parking lot energy management and has potential for practical smart grid scheduling applications.