1.河海大学能源与电气学院,江苏省南京市 211100;2.国电南瑞科技股份有限公司,江苏省南京市 211106
交直流配电网作为主动配电网的重要发展方向,协调其有功-无功资源以改善系统运行特性具有重要研究价值。首先,以基于多端柔性多状态开关(SOP)的交直流配电网为研究对象,计及储能、柔性负荷、SOP、光伏逆变器等关键设备运行约束,以及削峰填谷和电压偏移约束,构建了以运行成本最小为目标的交直流配电网有功-无功调度模型。然后,为适应光伏出力的随机性,构建了数据驱动的两阶段随机-分布鲁棒优化模型,并基于列与约束生成算法进行求解。最后,采用72节点和38节点交直流配电网作为算例,验证了模型的有效性。
国家自然科学基金资助项目(U1966205);中国科协青年人才托举工程资助项目(2021QNRC001)。
马智刚(1998—),男,硕士研究生,主要研究方向:交直流混合配电网优化运行。E-mail:mzg2108@hhu.edu.cn
卫志农(1962—),男,通信作者,教授,博士生导师,主要研究方向:电力系统运行分析与控制、输配电系统自动化。E-mail:wzn_nj@263.net
陈胜(1990—),男,博士,教授,硕士生导师,主要研究方向:市场环境下的综合能源系统规划与运行。E-mail:chensheng2019@hhu.edu.cn
1.College of Energy and Electrical Engineering, Hohai University, Nanjing 211100, China;2.NARI Technology Co., Ltd., Nanjing 211106, China
As an important development direction of the active distribution network, it is of significant research value to coordinate the active-reactive power resources of the AC/DC distribution network for improving the characteristics of system operation. First, taking the AC/DC distribution network based on multi-terminal soft open point (SOP) as the research object, an active-reactive power dispatch model of the AC/DC distribution network is constructed with the objective of minimum operation cost considering the key equipment operation constraints for the energy storage, flexible load, SOP, photovoltaic inverter, and the constraints of peak-shaving and valley-filling and voltage offset. Secondly, in order to adapt to the uncertainty of photovoltaic output, a data-driven two-stage stochastic-distributionally robust optimization model is constructed. Then, the model is solved based on the column and constraint generation algorithm. Finally, the 72-bus and 38-bus AC/DC distribution networks are used to verify the effectiveness of the proposed model.
[1] | 马智刚,卫志农,陈胜,等.基于柔性多状态开关的交直流配电网有功-无功优化调度[J].电力系统自动化,2023,47(6):48-58. DOI:10.7500/AEPS20220702002. MA Zhigang, WEI Zhinong, CHEN Sheng, et al. Active-Reactive Power Optimal Dispatch of AC/DC Distribution Network Based on Soft Open Point[J]. Automation of Electric Power Systems, 2023, 47(6):48-58. DOI:10.7500/AEPS20220702002. |