In anomaly detection model, the nearer and more frequent patterns are matched previously, and powered Euclid distances is used to account the similarity, and the anomaly data is sent to data analysis and response model.