Skip to main content

Table 4 Proposed frameworks of anomaly detection for electricity consumption

From: Anomaly detection for electricity consumption in cloud computing: framework, methods, applications, and challenges

Reference

Purpose

References

Rajasegarar et al. [57]

Discover unusual events in real-time

∙ It could be a predictor before anomalies happening.

  

∙ It might identify multilevel anomalies.

Chou and Telaga [58]

Assess the energy consumption for discovering potential power

∙ It was designed two schemes to identify energy theft attacks and faulty meters.

 

theft

∙ NTLs’ detection precision was improved and false positives were reduced.

  

∙ Technical losses also could be estimated.

Liu et al. [59]

Detect electricity theft and discover consumers involved

∙ Energy theft detection model can work for both dependent and independent data.

  

∙ The predictor variables were uncorrelated unless power theft occurs.

Arayaa et al. [60]

Describe the patterns of consumers’ electricity consumption

∙ Users were mapped into the 2D plane through PCA for showing data and detecting anomalies.

  

∙ Grid processing technology was used.

Xu et al. [61]

More accuracy of energy theft detection

∙ An information fusion method was used to discover energy theft attempts.

  

∙ Data mining techniques were utilized to detect anomalies by non-intrusive load monitoring.

  

∙A practical household load simulator was created to assess diverse techniques.