This study explores optimal energy management in smart buildings designed as microgrid systems, where the growing use of renewable energy and electric mobility creates challenges due to uncertainty and variability in supply and demand. The main goal is to reduce operational costs, cut emissions, and improve energy flexibility under uncertain conditions. To achieve this, a stochastic optimization framework is proposed, using Monte Carlo simulations to generate scenarios and k-means clustering to reduce their number efficiently. The smart building system includes solar panels, wind turbines, battery storage units, and two-way charging stations for electric vehicles and electric bicycles. At the core of the optimization lies the Snow Geese Algorithm (SGA), a nature-inspired metaheuristic that effectively handles the multi-objective nature of the problem. The key contribution of this work is the full integration of flexible mobility-based storage and demand response strategies into a smart building system modeled as a self-sufficient microgrid. Its originality lies in combining mobility-aware energy storage with a stochastic multi-objective structure, specifically designed for building-scale applications. This framework offers valuable insights into how smart buildings can achieve sustainable, reliable, and scalable energy performance. The method is tested in three scenarios: without mobility storage, with integrated electric mobility, and under uncertainty. Results show that integrating mobility storage reduces costs by 17.09% and emissions by 12.71%. When uncertainty is considered, operational costs rise by 17.33% and demand response expenses by 34.43%, though system flexibility remains strong. These findings emphasize the importance of integrating electric mobility and managing uncertainty in the next generation of energy-smart buildings.

中文翻译：

---

具有电动移动性和灵活性增强的智能建筑微电网的最优随机能源管理

本研究探讨了设计为微电网系统的智能建筑中的最佳能源管理，其中可再生能源和电动汽车的日益使用由于供需的不确定性和可变性而带来了挑战。主要目标是在不确定的条件下降低运营成本、减少排放并提高能源灵活性。为了实现这一目标，提出了一个随机优化框架，使用 Monte Carlo 模拟生成场景，并使用 k-means 聚类来有效地减少它们的数量。智能建筑系统包括太阳能电池板、风力涡轮机、电池存储单元以及电动汽车和电动自行车的双向充电站。优化的核心是雪雁算法 （SGA），这是一种受自然启发的元启发式算法，可有效处理问题的多目标性质。这项工作的主要贡献是将灵活的基于移动性的存储和需求响应策略完全集成到一个智能建筑系统中，该系统被建模为一个自给自足的微电网。它的独创性在于将移动感知储能与专为建筑规模应用设计的随机多目标结构相结合。该框架为智能建筑如何实现可持续、可靠和可扩展的能源性能提供了宝贵的见解。该方法在三种情况下进行了测试：没有移动存储、有集成电动汽车和在不确定性下。结果表明，集成移动存储可降低成本 17.09%，排放量降低 12.71%。如果考虑到不确定性，运营成本会增加 17.33%，需求响应费用会增加 34.43%，但系统灵活性仍然很强。 这些发现强调了在下一代能源智能建筑中整合电动汽车和管理不确定性的重要性。