考虑无功补偿的10 kV馈线拓扑数据驱动辨识方法

A Data-driven Approach for 10 kV Feeder Topology Identification Considering Reactive Power Compensation

中压配电网的拓扑结构对其运行分析与优化有着重要作用。近年来,配电网中的量测装置记录了大量稳态数据,如何利用量测数据辨识中压配电网拓扑结构成为研究重点。针对现有拓扑辨识方法未考虑馈线有无功补偿的情况,提出了考虑无功补偿的10 kV馈线拓扑辨识方法。首先,揭示了相邻节点电压降与节点间线路传输的有功功率和无功功率呈线性关系;然后,采用量测数据驱动的方式,通过线性回归误差平方和的比较,辨识10 kV馈线的末端节点和上级节点,进而通过迭代不断去除末端节点,直至剩余节点数为1;最后,利用迭代过程中产生的末端节点及上级节点信息重构馈线拓扑。算例验证了所提方法的有效性和鲁棒性。

The topology of medium-voltage distribution network plays an important role in its operation analysis and optimization. In recent years, a large amount of steady-state data has been recorded by measuring devices in distribution networks, and how to use the data to identify the topology of medium-voltage distribution network has become a research focus. Because the existing topology identification methods do not consider the reactive power compensation in the feeder, this paper proposes a data-driven approach for 10 kV feeder topology identification considering reactive power compensation. Firstly, the linear relationship between the voltage drop of adjacent nodes and the transmitted active and reactive power between nodes is revealed. Then, by comparing the square sum of the linear regression error through a data-driven approach, the end node and its upstream node of the 10 kV feeder are identified. By removing the end node, the identification can be performed again until only one node remains. Finally, the feeder topology is reconstructed with the information of the end node and its upstream node identified during the above iteration. Case studies show that the proposed method is effective and robust.