Credibility Assessment of Low-voltage Topology Changes Detection Based on Data Quality Analysis

A clear low-voltage topological relationship is the basis of low-voltage DC operation and maintenance. Without prior knowledge of the topology, the accuracy of the topological change detection method is hard to be quantified. Therefore, from the data quality perspective, a credibility assessment method of the data-driven low-voltage topology changes detection is proposed. Firstly, the influence mechanism of data quality on each link of the topological change detection method is analyzed, which provides the basis for the subsequent index system and evaluation method. Secondly, from the three dimensions of data missing, data error and data difference, a multi-dimensional and multi-level station data quality evaluation index system including 6 sub-items such as the proportion of missing measurement points is constructed to quantify data quality characteristics. Then, the influence degree and weight value of each index were clarified through sensitivity analysis, and the credibility assessment model is proposed to evaluate the reliability of the perception results of the topological changes of the station. Finally, the proposed assessment method is verified on the low-voltage station area of 15 users as an example.