基于Q学习的分布式电源选址定容
Distributed Generation Siting and Capacitation Based on Q-learning
西藏电网网架结构薄弱,供电半径大,使得对偏远地区的供电极不稳定。但其太阳能和水资源丰富,合理利用光伏和水电等地分布式电源(distributed generation, DG)为实现偏远地区用户平稳供电提供了可能。本文以西藏某地区电网实际运行新能源场站为研究对象,提出了西藏地区主动配电网中分布式电源的选址和容量模型,目标是使包括投资、操作和维护成本在内的总成本最小化。本文采用MATLAB平台和MATPOWER求解器进行求解。对负荷和分布式电源的不确定性进行了情景分析。为了分析非网络方案对最终规划结果和总投资的影响。通过典型IEEE 33实例验证了所提出的方法,比较分析了考虑主动管理和不考虑主动管理的规划结果,以及考虑环境效益和不考虑环境效益的规划结果,最终得到基于Q学习的分布式电源选址定容规划方案。为西藏地区后续的分布式电源系统建设提供了一些借鉴经验。
Distributed generation plays an important role in environmental issues and sustainable development worldwide. In this paper, a site selection and capacity model for distributed power within the active distribution networks is proposed. The goal is to minimize the overall costs, including investment, operation and maintenance costs. In this paper, the research employs the MATLAB platform and MATPOWER solver for solution and conducts scenario analyses to account for uncertainties in load and distributed power. In order to assess the impact of non-network scenarios on the final planning results and total investment, the study utilizes a typical IEEE 33 case for validation. The comparison of planning outcomes considering active management versus those without it, as well as considering environmental benefits versus disregarding them, is performed. Ultimately, the study presents a distributed power generation site selection and capacity planning scheme based on Q-learning, providing valuable insights for the development of distributed power systems in the Tibet region.
Q学习 / 配电网 / 分布式电源 / 选址定容 / 强化学习
Q-learning / distribution network / distributed power supply / site selection and capacity / reinforcement learning
| [1] |
|
| [2] |
丁明,王伟胜,王秀丽 |
| [3] |
张沈习,李珂,程浩忠 |
| [4] |
李振坤,陈思宇,符杨 |
| [5] |
李亮,唐巍,白牧可 |
| [6] |
李珂,邰能灵,张沈习 |
| [7] |
ZhangS,ChengH, |
| [8] |
郑乐,胡伟,陆秋瑜 |
| [9] |
|
| [10] |
丁明,刘先放,毕锐 |
| [11] |
吴小刚,刘宗歧,田立亭, |
| [12] |
赵波,肖传亮,徐琛 |
| [13] |
向育鹏,卫志农,孙国强 |
/
| 〈 |
|
〉 |