Multilevel and Synergistic Management Strategy for Power Quality Based on Data Driven

With the expansion of power grid scale, abnormal disturbance of power quality will bring serious harm. In view of the failure of traditional power management to effectively explore the potential capability of multiple power electronic equipment, a multilevel and synergistic management strategy for power quality based on data driven is proposed in this paper. Firstly, based on the local operation characteristics of the power quality control device, the important features in the power quality time series of each node of the distribution network are extracted. In order to improve the data quality, the data cleaning method based on k-nearest neighbor is adopted to construct isochronous dimension sequence array, and the correlation of each node data is analyzed by improved gray relational analysis (IGRA). The nodes are sorted according to the power quality correlation degree and the dominant nodes are determined by k-means++. The global governance under the power quality disturbance is realized through the regional regulation. Finally, the IEEE14-node simulation example is used to verify the accuracy of the proposed method and the rationality of improving the power supply quality in the whole network.