Evaluation of Fault Level of Sensitive Equipment Caused by Voltage Sag via Data Mining

So far, there is no literature to evaluate the fault level of sensitive equipment (FLSE) caused by voltage sag from the perspective of the power grid. In practice, although the FLSE are dominated by the voltage sag of node, the voltage sag of the node is associated with whole power grid, including the voltage grade and location of the node, the distance between the concerned node and location of fault, and the weather and date at the time of fault, which are all named as voltage sag properties (VSP). In view of this gap, this paper evaluates the FLSE using the long-time monitoring data of VSP of some regional power grid by data mining method based on multidimensional matrix simplification and the improved gray target theory. The data mining process can be divided into three steps: 1) construct a data base consisted of VSP and FLSE using long-time monitoring data, 2) mine the association rules between VSP and FLSE by multidimensional matrix simplification, and 3) match some actual scenario with the mined association rules by the improved gray target theory. Performance and effectiveness of the proposed method are verified through the simulation and field case.