A real-time energy consumption anomaly detection method combined with the structural characteristics of university buildings

Abstract

The invention belongs to the technical field of real-time anomaly detection, and specifically relates to a real-time energy consumption anomaly detection method combined with the structural characteristics of college buildings. The present invention adopts the SA-DBSCAN algorithm to adaptively identify the energy consumption mode of the building, and uses the C4.5 algorithm to construct the decision tree of the energy consumption mode; if there is no machine failure in the acquisition sub-node, it will be processed first according to the change of the parent-child structure of the building node Real-time incoming energy consumption, and then use the LOF algorithm to detect the abnormality after obtaining the category of the real-time energy consumption according to the decision tree; if the abnormality is judged, the decision tree is adjusted according to whether the boundary threshold is reached; Incremental updates to the building's energy consumption patterns and dynamically adjust the anomaly detection model based on whether the category of energy consumption patterns changes. The invention can effectively detect abnormal energy consumption and dynamically adjust the detection model to reduce misjudgment.

Description

technical field [0001] The invention belongs to the technical field of real-time anomaly detection, and in particular relates to a real-time energy consumption anomaly detection method combined with structural features of college buildings. Background technique [0002] Many universities have established energy consumption monitoring platforms and have collected a large amount of energy consumption data, which contains important information during building operation. In particular, real-time abnormal detection of energy consumption is conducive to discovering unreasonable aspects of building use and management, and timely adjustment can achieve the purpose of energy saving. At present, researchers have proposed many intelligent energy consumption anomaly detection methods, but they all use a fixed model for static analysis without considering environmental changes. The energy consumption environment of university buildings is relatively dynamic, and there are many types of ...

AI Technical Summary

Problems solved by technology

The addition and deletion of smart meters is the process of continuous improvement of the energy-saving supervision platform, and the practical significance of detecting such anomalies is not great

Moreover, this kind of situation is relatively common, and the detection of mixed together increases the difficulty of analysis and interpretation of other abnormalities, so it should be taken into consideration

Some smart meter nodes under the building node sometimes have empty values. This situation is likely to be caused by a machine failure in the collection machine of the smart meter

Images