The invention discloses a detection method of abnormal data in real-time data of a multi-element power network. The method includes: obtaining the real-time data of the multi-element power network tobe detected from the electric power real-time data, and establishing a data storage array; arranging the data to be detected according to the time series; introducing a detection interval and storingthe data to be detected into the detection interval; calculating a deviation distance of the data in the detection interval r, and calculating a relative deviation distance of the data according to the deviation distance; calculating a relative deviation distance of the data in the detection interval R and comparing the relative deviation distance threshold gamma i; By comparing the relative deviation distance with the relative deviation distance threshold gamma i, it is judged whether the detected data is abnormal data or not, and the abnormal data is stored in the abnormal database. It is easy to realize, and can quickly and accurately complete the anomaly detection of massive real-time electric power data, timely discover the anomaly data in the multi-grid, and ensure the safety of themulti-grid data.