Robust design and operation of stand-alone hybrid renewable energy systems

混合可再生能源系统（HRES）结合了传统和可再生能源发电机和能源存储，正在成为电网连接的可行和有吸引力的替代方案，不仅为偏远和孤立的社区服务，也为独立的建筑和城市社区服务。然而，考虑到可再生能源的间歇性和不可预测性、需求的可变性和能源储存的高成本，它们需要仔细设计和操作，以便可靠、经济和环保地满足需求。目前的HRES规划方法要么过于乐观，天真地假设供给和需求是确定的，或者有已知的概率分布，要么过于悲观，根据最坏的情况进行规划，导致系统过大，在技术/经济上不可行。在本研究计划中，我们探讨了新的实用和平衡的方法，即既不乐观也不悲观，有效地利用历史供需数据中包含的信息来捕获和处理不确定性问题。我们将在不确定条件下应用和扩展现代数学和统计技术来解决HRES规划问题，包括时间序列分析、鲁棒模拟优化、数据聚类和分布鲁棒优化。该研究还将为这些大规模和具有计算挑战性的问题探索新的有效解决方法，这些方法也可以应用于处理其他决策问题。此外，我们将研究使用热电联产（CHP）装置的建筑物和社区的电力和热能供应的协同优化，同时考虑电动汽车和太阳能屋顶等新技术的影响。这项研究有望通过更好的设计和操作，显著提高单机HRES的技术和经济可行性，使其与我们目前基于化石的集中式能源供应系统相当（甚至更好）。一方面，这将使加拿大北部的农村社区，包括土著社区，拥有可靠和经济的能源供应，利用当地丰富的可再生能源，如风能、水能、潮汐能和太阳能。另一方面，在并网城市和郊区社区的能源管理系统中实施拟议的强有力的HRES规划框架，将增加可再生能源的渗透，并促进向更可持续和更安全的能源系统过渡。

Hybrid renewable energy systems (HRES), which combine conventional and renewable energy generators and energy storage, are becoming a viable and attractive alternative to grid connection for serving not only remote and isolated communities, but also independent buildings and urban communities. However, given the intermittent and unpredictable nature of renewable energy sources, the variability of demand and the high cost of energy storage, they need to be carefully designed and operated so they can serve demand reliably, economically and eco-friendly. Current HRES planning approaches are either too optimistic, naively assuming that the supply and demand are deterministic or have known probability distributions, or too pessimistic, planning them based on the worst-case scenario which leads to oversized and technically/economically infeasible systems. In this research program, we investigate new practical and balanced, i.e., neither optimistic nor pessimistic, approaches for planning HRES that effectively utilize the information contained in the historical supply and demand data to capture and deal with the uncertainty issue. We will implement and extend modern mathematical and statistical techniques in decision-making under uncertainty to tackle the HRES planning problem, which include time-series analysis, robust simulation-optimization, data clustering and distributionally-robust optimization. The research will also explore new efficient solution methods for these large-scale and computationally-challenging problems that can be applied also to handle other decision-making problems. Furthermore, we will study the co-optimization of electrical and thermal energy supply in buildings and communities that use combined-heat-and-power (CHP) units while considering the effect of new technologies such as electric vehicles and solar roofs. This research is expected to significantly improve the technical and economical feasibility of stand-alone HRES through better design and operation, putting them on par with (or even better than) our current fossil-based centralized energy supply systems. On one hand, this will enable rural Northern communities in Canada, including aboriginal communities, to have reliable and economical energy supply that uses locally-abundant renewable energy resources such as wind, hydro, tidal and solar energies. On the other hand, implementing the proposed robust HRES planning framework in energy management systems for grid-connected urban and suburban communities will increase the penetration of renewable energy sources and foster the transition towards a more sustainable and secure energy system.