Rule‐based classification of energy theft and anomalies in consumers load demand profile

Abstract

The invent of advanced metering infrastructure (AMI) opens the door for a comprehensive analysis of consumers consumption patterns including energy theft studies, which were not possible beforehand. This study proposes a fraud detection methodology using data mining techniques such as hierarchical clustering and decision tree classification to identify abnormalities in consumer consumption patterns and further classify the abnormality type into the anomaly, fraud, high or low power consumption based on rule-based learning. The proposed algorithm uses real-time dataset of Nana Kajaliyala village, Gujarat, India. The focus has been on generalizing the algorithm for varied practical cases to make it adaptive towards nonmalicious changes in consumer profile. Simultaneously, this study proposes a novel validation technique used for validation, which utilizes predicted profiles to ensure accurate bifurcation between anomaly and theft targets. The result exhibits high detection ratio and low false-positive ratio due to the application of appropriate validation block. The proposed methodology is also investigated from point of view of privacy preservation and is found to be relatively secure owing to low-sampling rates, minimal usage of metadata and communication layer. The proposed algorithm has an edge over state-of-the-art theft detection algorithms in detection accuracy and robustness towards outliers.

Additive Ensemble Neural Network with Constrained Weighted Quantile Loss for Probabilistic Electric-Load Forecasting

“…In these alternative approaches, the application of accurate forecasting models is crucial. The detection of abnormalities in load consumption patterns to identify energy theft or other types of attacks is studied in [ 58 ], based on a hierarchical clustering and decision trees classification. A similar approach is presented in [ 59 ], which also uses a decision tree algorithm without prior clustering.…”

A novel unsupervised feature‐based approach for electricity theft detection using robust PCA and outlier removal clustering algorithm

“…In Reference 11, power theft consumers were identified by formulating a number of rules that were based on conditional probability and K‐means clustering techniques. In another study, 12 the authors developed an NTL detection model based on the amalgamation of hierarchical clustering and decision trees technique to identify the abnormalities in the user's consumption data. Finally, a rule‐based learning approach was adopted to classify the consumers based on the severity of the fraud.…”

Time-Pattern Profiling from Smart Meter Data to Detect Outliers in Energy Consumption

“…Information can be mined and employed for decision-making processes that ultimately allow for the optimisation of the infrastructure management. However, a significant amount of research also exists (some of which is presented in this paper) on how the data collected from the smart meters can be used to detect energy usage patterns in residential homes via user profiling techniques [1]. There are close synergies between the level of fuel consumption and the behaviour of the occupants within their homes [2].…”

“…IoT 2020, 1 Analysing the consumption patterns collected by smart meters at frequent intervals using advanced data analysis techniques, provides a practical solution as a smart home Internet of Things (IoT) application for modelling behavioural patterns. For example, Amri et al used k-means to provide insight into seasonal consumption patterns [3].…”