李雪晴, 艾欣, 王昊洋. 基于改进分析目标级联法的交直流混联配电网分布式优化运行[J]. 现代电力. DOI: 10.19725/j.cnki.1007-2322.2022.0271
引用本文: 李雪晴, 艾欣, 王昊洋. 基于改进分析目标级联法的交直流混联配电网分布式优化运行[J]. 现代电力. DOI: 10.19725/j.cnki.1007-2322.2022.0271
LI Xueqing, AI Xin, WANG Haoyang. Distributed Optimization Operation of AC/DC Hybrid Distribution Network Based on Improved Analytical Target Cascading Method[J]. Modern Electric Power. DOI: 10.19725/j.cnki.1007-2322.2022.0271
Citation: LI Xueqing, AI Xin, WANG Haoyang. Distributed Optimization Operation of AC/DC Hybrid Distribution Network Based on Improved Analytical Target Cascading Method[J]. Modern Electric Power. DOI: 10.19725/j.cnki.1007-2322.2022.0271

基于改进分析目标级联法的交直流混联配电网分布式优化运行

Distributed Optimization Operation of AC/DC Hybrid Distribution Network Based on Improved Analytical Target Cascading Method

  • 摘要: 交直流混联的配电方式是未来配电网的一种重要发展方向,为克服其中各子系统分属不同利益主体而带来的功率信息传递的限制,提出了一种基于改进分析目标级联法的交直流混联配电网分布式优化运行方法。首先,针对交直流混联配电网的直流、换流站和交流区域,分别建立包含自身区域运行优化与区域间交互功率运行优化的模型;其次,采用分布式优化的思想,将交互功率作为共享变量,对模型目标函数进行解耦,并在传统的分析目标级联法中引入一个平衡系数,来解决因惩罚乘数初值选择不当,子系统函数项与惩罚项权重不平衡情况下算法性能不好的问题,在一致性约束下迭代求解分布式模型;最后,通过算例仿真,进行模型的有效性、改进算法性能以及分布式模型通用性的验证。

     

    Abstract: AC/DC hybrid distribution mode is an important development direction of future distribution network. In order to overcome the limitation of power information transmission caused by different interest subjects of each subsystem, a distributed optimization operation method of AC/DC hybrid distribution network based on improved analytical target cascading method was proposed. Firstly, in allusion to DC, VSC and AC region of AC/DC hybrid distribution network, the optimization models including regional operation optimization and inter-regional interactive power operation optimization were established respectively. Secondly, using the idea of distributed optimization, the interaction power was used as a shared variable, the model objective function was decoupled, and a balance coefficient was introduced in the traditional analytical target cascading method to solve the problem of poor algorithm performance due to improper selection of the initial value of the penalty multiplier and unbalanced weights of subsystem function terms and penalty terms, then the distributed model was solved iteratively under the consistency constraint. Finally, the effectiveness of the improved algorithm and the universality of the distributed model are verified by example simulation.

     

/

返回文章
返回