近年来，电动汽车、分布式发电、分布式储能、电采暖/制冷等分布式资源发展迅速，为电力系统提供了宝贵的调节能力。聚合海量分布式资源参与电力系统的运行调控，是促进可再生能源消纳，实现能源系统清洁化转型的有力手段。然而，由于分布式资源规模庞大且种类繁多，且其调节能力受到配网安全限制的制约以及不确定性的影响，它们提供的调节潜力难以在电力系统的实际运行调控中得到充分的利用。本文针对大规模分布式资源参与电网运行的调节能力聚合建模与优化调度开展了研究，主要工作如下： (1) 从分布式资源的单体调节能力模型出发，通过严格的理论推导得到了精确聚合模型解析表达式。随后给出了精确聚合模型的多种形式，分析了其物理意义以及理论性质，作为大规模分布式资源参与电网运行的理论基础。 (2) 将精确聚合模型的性质应用到近似模型的构建过程中，提出了一系列高效的近似聚合建模方法，得到外部与内部近似的聚合调节能力模型。所提的近似聚合方法相比文献中的方法实现了精度和计算效率的显著提升，使得调节能力聚合建模技术得以满足电力系统实际应用的需求。 (3) 提出了配网层的分布式资源聚合建模与优化调度方法。一方面，将配网安全约束引入到分布式资源聚合过程中，推导出了主配网接口处的调节能力模型形式，并提出了适用于不同应用场景的多种参数计算方法，算例仿真验证了所提模型与方法的正确性与性能优势。另一方面，提出了一种与调节能力模型相匹配的成本与价值统一量化方法，并在此基础上设计了聚合分布式资源参与大电网运行调控的配网层优化调度模型与收入分配策略。 (4) 提出了基于调节能力聚合的输配协同优化调度方法。针对机组组合与经济调度两个场景，将聚合调节能力模型应用到输电网级的运行调控中，在调动大规模分布式资源提升输配系统整体经济性与新能源消纳能力的同时仍然保持了高计算效率。提出了基于双层迭代的分布鲁棒机会约束近似方法，用于处理调度问题中聚合调节能力的不确定性，克服了传统方法存在的过保守乃至不可行的缺点。 本文的研究成果可为大规模分布式资源参与电力系统运行提供理论基础和实践路径参考，促进电力系统充分挖掘分布式资源的调节潜力，推动新型电力系统构建和能源结构转型的进程。

The proliferation of distributed energy resources (DERs) such as electric vehicles, distributed generation, distributed energy storage systems, and thermostatically controlled loads can provide huge dispatchable potential for power systems. Activating the flexibility of numerous DERs in power system operations can promote renewable energy integration and clean energy transition. However, DERs have a very large population, and their flexibility is uncertain and restricted by the distribution system security constraints, so their flexibility is often underutilized in the actual operation and control of power systems. This thesis focuses on the modeling of flexibility aggregation and optimal dispatch of large-scale DERs in power system operations, with the main contributions summarized as follows:(1) Starting from an individual flexibility model of DERs, the explicit mathematical expression of the exact flexibility aggregation model is derived based on rigorous theoretical analysis. Various forms of the exact aggregation model are then presented, and its physical meaning and theoretical properties are analyzed. The exact aggregation model and its theoretical properties can serve as the theoretical foundation for large-scale DERs participating in power system operations.(2) Outer and inner approximate flexibility aggregation modeling methods are proposed based on the properties of the exact aggregation model. Compared to the approximate modeling methods in the literature, the proposed methods show significant improvements in accuracy and computational efficiency, enabling the flexibility aggregation technique to meet the requirements of practical power system applications.(3) A flexibility aggregation method and an optimal dispatch model at the distribution system level are proposed. First, the flexibility model at the transmission-distribution interface, which is based on the aggregation of all the flexible DERs while considering the distribution system security constraints, is proposed, along with effective parameter calculation methods. Numerical simulations validate the correctness and advantages of the proposed models and methods. Second, a unified flexibility cost and value quantification method aligned with flexibility models is proposed, based on which an optimal dispatch model and a revenue allocation strategy for activating the DERs in distribution systems to participate in the transmission system operation are designed.(4) A transmission-distribution coordinated dispatch framework based on flexibility aggregation is proposed. For unit commitment and economic dispatch scenarios, the aggregated flexibility models of distribution systems are applied to the optimal dispatch at the transmission level, enhancing the renewable energy consumption and the overall economic efficiency of the integrated transmission-distribution system while maintaining high computational efficiency. A bi-level iterative approximation method for distributionally robust joint chance constraints is proposed, which is utilized to handle the uncertainty of aggregated flexibility in the dispatch problem, overcoming the over-conservativeness and potential infeasibility limitations of traditional Conditional-Value-at-Risk-based methods.The contributions of this thesis provide a theoretical foundation and practical reference for the participation of large-scale DERs in power system operations, promoting the DERs' flexibility exploitation, the new power system construction, and the clean energy transition.