Power system bad load data detection based on an improved fuzzy C-means clustering algorithm

Massive electricity information are measured, and lots of bad data still exist. The mathematical method of hill-climbing to determine the initial clustering centers and clustering number was studied in this paper, which overcomes the subjectivity of determining clustering number and prevents the objective function from falling into local minimum. Then a bad data detection method is proposed, which combines the fuzzy C-means (FCM) clustering algorithm with hill-climbing method to determine the number of clusters and initial clustering centers in advance, cluster the load data, and finally generate a feasible region matrix by extracting the characteristic curve. At last, some cast studies are provided to demonstrate the effectiveness of the proposed method.