Advanced Distributed Decision-making framework for Smart grid aggregators

今天，配电网正处于从需求驱动型向主动资产驱动型转变的时代，其特征是扩大分散发电机组的措施，彻底重新思考电力市场，以及终端用户越来越多地参与需求响应(DR)计划。为此，分销网络将从自上而下的结构演变为自下而上的结构，客户在分销系统中的角色将从被动参与转变为主动参与。新能源市场参与者，如联网微电网、主动建筑聚合器和负荷聚合器，可以通过其本地发电、存储和需求灵活性为配电网络运营商提供支持。在聚合器内运行这种玩家集群可能是利用分布式能源(DER)、向批发电力系统提供低成本可靠性产品、即向传输系统提供辅助服务的最佳方式。能源市场对聚合或需求响应聚合很感兴趣，因为这种结构可以实施强大的经济动机来吸引许多参与者。随着这种转变，运营商通过适当协调现有资源和需求侧管理机制，继续保持电网中的平衡至关重要。 尽管有这些优势，但与容纳大量参与者(微电网、负荷、DER等)相关的挑战仍然存在。其中一些主要挑战是：1)集中式优化方法的困难；2)行为及其不同利益的异质性；3)调用聚集器时负载和分布式能源资源可用性的不确定性；以及4)不同客户之间需求响应的回报的公平性。在这个发现计划中，我们的目标是克服与许多需求响应参与者的协调相关的一些挑战，特别是那些与缓解维度诅咒相关的挑战，从而允许在聚合层面对冲不确定性。我的研究计划的长期目标是开发先进和高性能的工具，对配电网络中的各种能源市场参与者进行建模和分布式优化。特别是，我们的目标是通过适当的方法适当地捕捉可再生能源和负载不确定性，在能源交易机制和参与者之间的DR奖励分配方面实现公平。拟议中的Discovery计划将有助于分布式能源的发展，以及智能电网中聚合器的增长。该提案将为一项长期计划奠定基础，确保：1)加拿大总部成为智能电网系统算法设计领域的世界领先者，2)对新的更大项目(如NSERC-Alliance)的重大影响力，这些项目由海德鲁-魁北克、施耐德等合作伙伴共同资助。

Today, electricity distribution networks are in the era of transforming from a demand-driven to an active asset driven entity, portrayed by expanding measures of decentralized generation units, complete rethinking of the electricity market and an increasing participation of end-users in demand response (DR) programs. To this end, distribution networks will be evolving from a top-down structure into a bottom-up structure and the role of customers in distribution systems will shift from passive to an active participation. New energy market players such as networked microgrids, active building aggregators and load aggregators can support the distribution network operator through their local generation, storage and demand flexibility. The operation of such clusters of players within the aggregator is possibly the best way to utilize distributed energy resources (DER), to supply low cost reliability products to the wholesale electricity system i.e., providing ancillary services to the transmission system. Aggregation or demand response aggregators are interesting to the energy market as such structure could implement powerful economic motivators to attract many participants. With such transformation, it is vital for the operators to continuously preserve the balance in the electricity grid through appropriate coordination of available resources and demand side management mechanisms. Despite these advantages, there exist several challenges associated with accommodating large number of participants (microgrids, loads, DERs, etc.). Some of the main challenges are: 1) intractability of centralised optimization approaches; 2) the heterogeneity of actions and their diverging interests; 3) the uncertainty in the availability of loads and distributed energy resources when called upon the aggregator and 4) the fairness in the rewards from demand response among various customers. In this discovery program, we aim to overcome some of the challenges associated with coordination of many demand response players, especially those related to mitigating the curse of dimensionality allowing to hedge against uncertainty at aggregation level. The long-term objective of my research program aims to develop advanced and high-performance tools to model and distributively optimize various energy market players in the distribution network. Particularly, we aim to capture properly renewable energy sources and load uncertainties via appropriate methods, achieve fairness in energy trading mechanisms and in DR rewards distribution among players. The proposed Discovery program will contribute to the development of distributed energy resources, the growth of aggregators in smart grids. The proposal will set the foundations for a long-term program ensuring that: 1) Canadian HQPs become world leaders in the field of algorithms design in smart grids systems and 2) Significant leverage for new larger projects such NSERC-Alliance co-financed by partners like Hydro-Quebec, Schneider.