| 引用本文: | 梁海平,王岩,刘英培,王鑫明.计及源荷不确定性的混合交直流主动配电网分层-分布式优化调度[J].电力自动化设备,2021,41(12): |
| LIANG Haiping,WANG Yan,LIU Yingpei,WANG Xinming.Hierarchical-distributed optimal scheduling of hybrid AC/DC active distribution network considering source and load uncertainties[J].Electric Power Automation Equipment,2021,41(12): |
|
| 摘要: |
| 混合交直流主动配电网是未来配电网发展的主要形式，针对混合交直流主动配电网集中式优化存在优化时间长、对区域内隐私保护性不强以及源荷出力不确定性的问题，提出计及源荷不确定性的混合交直流主动配电网分层-分布式优化调度策略。该策略在混合交直流主动配电网区域内以预测场景和抽样场景的期望运行成本之和最小为目标，建立两阶段随机规划模型，采用Benders分解算法分解成主、子问题交替迭代求解；在区域间为兼顾各区域数据私密性及实现并行计算，以各自区域内主问题运行成本最优为目标，采用同步型交替方向乘子法进行分布式计算。所提策略通过各区域主问题连接区域间和区域内形成整体优化，从而保证模型达到全局最优。最后，通过算例分析验证了所提的分层-分布式优化调度策略的有效性。 |
| 关键词: 混合交直流主动配电网 两阶段随机规划 Benders分解算法 同步型交替方向乘子法 分层-分布式优化调度 |
| DOI：10.16081/j.epae.202107003 |
| 分类号:TM727;TM73 |
| 基金项目:国家自然科学基金资助项目(51607069) |
|
| Hierarchical-distributed optimal scheduling of hybrid AC/DC active distribution network considering source and load uncertainties |
|
LIANG Haiping1, WANG Yan1, LIU Yingpei1, WANG Xinming2
|
|
1.School of Electrical and Electronic Engineering, North China Electric Power University, Baoding 071003, China;2.State Grid Hebei Electric Power Co.,Ltd.,Shijiazhuang 050021, China
|
| Abstract: |
| Hybrid AC/DC active distribution network represents the main development form of future distribution network. Aiming at the problems of huge computational burden, weak protection of local privacy and both source and load output uncertainties of centralized optimization for hybrid AC/DC active distribution network, a hierarchical-distributed optimal scheduling strategy of hybrid AC/DC active distribution network conside-ring the source and load uncertainties is proposed, which aims to minimize the sum of expected operating cost under both predicted and sample scenarios in the region of hybrid AC/DC active distribution network. A two-stage stochastic programming model is established in each region, and the Benders decomposition algorithm is employed to solve the main and sub problems iteratively. To protect the data privacy of each region and accomplish parallel computing, the model aims to optimize the operating cost of main problem in each region, and the synchronous alternating direction method of multipliers is utilized for the distributed calculation. In this strategy, the main problem of each region is interrelated by the inter-and intra-regional optimization problems to ensure globally optimal solutions. Finally, the case analysis validates the effectiveness of the proposed hierarchical-distributed optimal scheduling strategy. |
| Key words: hybrid AC/DC active distribution network two-stage stochastic programming Benders decomposition algorithm synchronous alternating direction method of multipliers hierarchical-distributed optimal schedu-ling |
