Data mining-based intelligent diagnosis method and diagnosis system for electricity utilization inspection

The invention discloses a data mining-based intelligent diagnosis method and diagnosis system for electricity utilization inspection. The diagnosis method comprises the following steps of: 1, acquiring electricity utilization information of a user; 2, preprocessing the data sources acquired in the step 1; 3, performing clustering analysis on the electricity utilization information data preprocessed in the step 2 by adopting an FCM clustering algorithm to obtain a typical load characteristic curve; 4, calculating the similarity between the user and the typical load characteristic curve, and screening abnormal power utilization suspected users; 5, constructing an abnormal power utilization evaluation index system; 6, constructing an anti-electricity-stealing behavior recognition model by using a fuzzy neural network; and 7, predicting the electricity larceny behavior by using the anti-electricity larceny behavior recognition model obtained in the step 6. The diagnosis system comprises a data layer, an applicati