博弈论视角的光伏用户群互动式能量优化方法

The cluster of photovoltaic (PV) prosumers is a typical form of large scale distributed PV power to connect to the distribution network in the future. Regulating the resources on the demand side through interactive energy management is one of the feasible approaches to promote the effective consumption and utilization of renewable energy in clusters. From the perspective of Game theory, the project intensively focuses on the method of interactive energy management of PV prosumer cluster. There exists three difficult problems: the construction of Game theory framework, the modeling of optimization and the scheduling strategies in multiple time scales. On the level of Game theory framework, the multi-agent Game framework such as non-cooperative Game, leader-follower Game and coalitional Game is established in the market-oriented business model according to the features, relationships and scheduling strategies of operators and PV consumers. On the level of optimization model, the optimization model of PV consumer cluster would be developed from the centralizing and autonomous manner to the multi-agent interaction one, and then establishes the distributed energy optimization model and its solution algorithm which contains multi-agent interest demands and grid constraints. On the level of scheduling strategies, it is the mechanism of interactive energy management that is established in multiple time scales, in which differentiated scheduling strategies and auxiliary mechanism under different scenarios are proposed as well. The research achievements will promote the market transactions and effective utilization of distributed PV at high proportion and support the development of energy management system on the electricity consumer side.

光伏用户群是未来分布式光伏电源大规模接入配电网所呈现的典型形态，通过互动式能量优化充分调动需求侧资源，是促进可再生能源高效消纳利用的可行途径之一。本课题重点研究博弈论视角下光伏用户群的互动式能量优化方法，以博弈论框架构建、优化问题分解建模、多时间尺度调度策略等三类难点问题作为研究的切入点。在博弈论框架层面，从市场化商业模式出发，根据运营商及光伏产消者的角色特性、利益关系与调控策略，构建非合作、主从、合作等多种形式的多主体博弈框架；在优化模型层面，将光伏用户群的优化运行问题由集中自律模式延伸至多主体互动模式，建立涵盖多主体利益诉求及电网约束条件的分布式能量优化模型及求解算法；在调度策略层面，建立多时间尺度的互动式能量管理调度机制，提出不同场景下的差异化调度策略与辅助机制。研究成果将促进高比例分布式光伏的市场交易与高效消纳利用，并支撑用户侧能量管理系统的研发与标准制定。

本课题的研究基本按照申请书的计划内容及时间周期执行，重点针对四类问题展开研究，包括：（1）基于非合作博弈架构的产消者能量调控方法；（2）基于合作博弈架构的产消者能量调控方法；（3）基于主从博弈架构的产消者能量调控方法；（4）考虑产消者能量优化及随机性的分布式调压/调频方法。针对第一类问题，以非合作博弈为前提，提出了基于模型驱动的产消者综合能量调控方法、基于数据驱动的主观产消者互动式能量共享方法，以及基于匹配博弈的产消者P2P能量共享方法。针对第二类问题，在合作博弈的框架下，提出了光伏用户群能量共享模式，并提出了基于谈判博弈的含光伏用户能源枢纽能量调控策略，还研究了对等模式下配电网与通信网能量协同优化方法。针对第三类问题，提出了基于展望理论的非理性产消者能量共享方法，研究了含产消者的多能源工业园区能量交互与管理方法，建立了多综合能源系统低碳经济调度和能量共享方法，提出了主从模式下含规模化5G基站的配电网优化运行策略。通过上述的系列化研究，积累了较为丰硕的研究成果，共计发表论文26篇，其中SCI论文17篇、EI论文23篇，申请发明专利10项，3项处于审理阶段。通过与电网企业及设备制造企业合作，相关理论成果也获得了较好的应用。通过本项目的研究，共计培养博士生3名，硕士生8名。