Power consumption data anomaly detection method and device, computer equipment and storage medium

Abstract

The invention relates to a power consumption data anomaly detection method and device, computer equipment and a storage medium. The method comprises the steps of obtaining a power consumption data sequence; inputting the electricity consumption data sequence into a pre-constructed electricity consumption prediction model to obtain electricity consumption prediction data; determining a difference value between the electricity consumption prediction data and the electricity consumption real data, and if the difference value is greater than a preset threshold value, identifying the electricity consumption real data as candidate abnormal data; when the candidate abnormal data reaches a preset abnormal condition, determining that the current power consumption data abnormal detection result is in an abnormal state. According to the invention, the power consumption data sequence is identified through the pre-constructed power consumption prediction model to obtain the power consumption prediction data, whether the prediction data is abnormal data is identified according to the difference value between the prediction data and the real data, and when the abnormal data reaches the preset condition, generation of the abnormal detection result is triggered, so that the effect of detecting the abnormal power consumption data without manually marking the features is achieved, and the power consumption data anomaly detection efficiency is improved.

Description

technical field [0001] The present application relates to the field of electric power technology, and in particular, to a method, device, computer equipment and storage medium for abnormality detection of power consumption data. Background technique [0002] With the development of machine learning theory, scholars and practitioners gradually apply machine learning technology to the task of anomaly detection of electricity consumption data. [0003] At present, due to the drift, fluctuation and other characteristics of electricity consumption data, it is more and more difficult to manually design feature models, and the time series characteristics of electricity consumption in different industries are very different. The scope of application is very limited, which leads to the low detection efficiency of the existing abnormal detection methods of electricity consumption data. [0004] Therefore, there is also a need for a method for detecting abnormality of power consumptio...

AI Technical Summary

Problems solved by technology

[0003] At present, due to the characteristics of drift and fluctuation in electricity consumption data, it is becoming more and more difficult to manually design feature models, and the timing characteristics of electricity consumption in different industries are very different. Unsupervised learning methods using a unified distance measurement method and similarity measurement method The scope of application is very limited, resulting in low detection efficiency of the existing electricity data anomaly detection methods

Images