用户用电负荷变化的异常检测与识别OA北大核心CSTPCD

In the era of smart grids,most abnormal electricity consumption behaviors will be accompanied by the changes in electricity load.The research on user electricity consumption behavior is crucial for the operation and management of the power system.On this basis,a method of directly using load data for calculation is proposed to determine whether users have abnormal electricity consumption behavior,which can be realized by the correlation between the characteristic electricity consumption load curve and the daily electricity consumption load curve.In the process of correlation calculation,the Euclidean distance is combined with the Pearson correlation coefficient to more accurately determine whether the user's electricity load has undergone significant changes.An adaptive threshold strategy is adopted for the experimental research on 500 sets of data to improve the accuracy and flexibility of judgment.The correlation calculation results show that 122 groups'data have excessive load changes,and 99 groups'data have abnormal load events.The accuracy of this method can reach 81.1%.