Research on data-driven active distribution network operation status evaluation method

Due to the imbalanced nature of active distribution network operation status data, it is difficult to control the deviation between its evaluation results and the actual situation. Therefore, a data-driven active distribution network operation status evaluation method is proposed. Using a data-driven type-2 fuzzy algorithm to determine the interval range of clustering centers for active distribution network operation data, and using the IT2FKM clustering method to constrain the shift of clustering centers from minority to majority clusters in the case of imbalanced clusters. In the state evaluation stage, multi-source fusion is performed on the clustered data to evaluate the specific operating status from a voltage perspective. In the test results, the evaluation results of the design method have the highest fitting degree with the actual values, and the corresponding error is also stable within 0.1 p.u. Compared with the control group, it has a significant advantage.