| 引用本文: | 孙国强,王善磊,陈胜,吴晨,胡国伟,周亦洲,卫志农.基于双层Kriging元模型算法的多产消代理商主从博弈能量管理模型[J].电力自动化设备,2021,41(11): |
| SUN Guoqiang,WANG Shanlei,CHEN Sheng,WU Chen,HU Guowei,ZHOU Yizhou,WEI Zhinong.Stackelberg game model for energy management of multiple prosumer aggregators based on bilevel Kriging meta model algorithm[J].Electric Power Automation Equipment,2021,41(11): |
|
| 摘要: |
| 分布式能源的蓬勃发展为建设低碳高效的能源系统提供了新的途径，但同时也带来用户侧行为随机、能量管理困难等难题。为实现用户侧低碳能源系统的柔性调控与高效管理，提出产消代理商的概念，并对其具体含义作了详细说明。在此基础上，为进一步实现用户侧清洁能源的就地消纳并促进多产消代理商间的能量共享，构建了市场运营商(MO)和多产消代理商间主从博弈模型。该模型中，MO作为博弈的领导者，通过电价优化引导产消代理商的购电/售电行为；而产消代理商作为博弈的跟随者，在收到MO制定的电价后，以最小化用电成本为目标对内部各聚合单元进行优化。同时，为克服下层模型中由布尔变量带来的求解困难问题，采用双层Kriging元模型算法实现博弈模型的求解，减少了下层模型的调用次数，显著提高了计算效率。算例验证了所构建模型的有效性。 |
| 关键词: 低碳能源 产消代理商 主从博弈模型 能量管理 双层Kriging元模型 |
| DOI：10.16081/j.epae.202109015 |
| 分类号:TM73 |
| 基金项目:国家自然科学基金资助项目(52077060)；国网江苏省电力有限公司科技项目(J2021143) |
|
| Stackelberg game model for energy management of multiple prosumer aggregators based on bilevel Kriging meta model algorithm |
|
SUN Guoqiang1, WANG Shanlei1, CHEN Sheng1, WU Chen1,2, HU Guowei2, ZHOU Yizhou1, WEI Zhinong1
|
|
1.College of Energy and Electrical Engineering, Hohai University, Nanjing 211100, China;2.Economic Research Institute of State Grid Jiangsu Electric Power Co.,Ltd.,Nanjing 210008, China
|
| Abstract: |
| The vigorous development of distributed energy provides a new way to build a low-carbon and efficient smart grid, but it also brings problems such as random behavior of users and difficulties in energy management. In order to realize the flexible regulation and efficient management of low-carbon energy system on the user side, the concept of prosumer aggregators is proposed, and its specific meaning is explained in detail. On this basis, the Stackelberg game model between MO(Market Operator) and multiple prosumer aggregators is built to further realize the local consumption of clean energy and promote the energy sharing. In this model, MO, as the leader of the game, guides prosumer aggregators to purchase/sell electricity through dynamic pricing, while prosumer aggregators, as the followers of the game, optimize the internal aggregation units to minimize the cost of electricity consumption after receiving the price set by MO. At the same time, in order to overcome the solving difficulty caused by Boolean variables in the lower-level model, the bilevel Kriging meta model algorithm is used to solve the game model, which reduces the call times of the lower-level model and significantly improves the computational efficiency. The effectiveness of the built model is verified by an example. |
| Key words: low-carbon energy prosumer aggregators Stackelberg game model energy management bilevel Kriging meta model |
