Power Quality Situation Awareness in Distribution Station Area Based on Data Mining

With the continuous development of renewable energy and the rapid growth of user power demand, it is significant how to obtain quickly and accurately the operation status and future development trend of power quality in the distribution station area. A data mining-based power quality situation awareness method in distribution station area is proposed. Firstly, in the stage of situation assessment, the criteria importance though intercriteria correlation(CRITIC) is used to mine the data of power quality acquisition device to determine the index weight at the time of on-line assessment, and the real-time data is used to construct the dynamic evaluation matrix of power quality ground to realize online evaluation of power quality; Then, in the stage of situation prediction, the complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN) is used on the basis of single long-short term memory (LSTM) to master its situation development; Finally, an example shows that the method can grasp the situation of power quality in distribution area in time, which is conducive to the realization of lean management.