Electricity Theft Detection in Smart Grid Systems: A CNN-LSTM Based Approach

Hasan, Md. Nazmul; Toma, Rafia Nishat; Nahid, Abdullah-Al; Islam, M. M. Manjurul; Kim, Jong‐Myon

doi:10.3390/en12173310

Cited by 283 publications

(236 citation statements)

References 55 publications

Supporting

Mentioning

236

Contrasting

Order By: Relevance

“…This imbalanced nature of the dataset adversely affects the performances of the supervised learning techniques because of the biasn towards the majority class. Hence, to reduce the class inequality problem, there are two major types of techniques, known as the cost function-based and sampling-based techniques [22].…”

Section: Data Balancingmentioning

confidence: 99%

“…For this reason, the energy consumption data are transformed into 2-D weekly data and serve as an input to the UNet model. Furthermore, the authors in [20,22,40] have used a traditional CNN technique to derive high-level features from electricity consumption data. However, if we use the regular convolution network with the pooling and dense operations, the model will only extract high-level features of "what," but not their localization information, "where."…”

Section: Unet Modulementioning

confidence: 99%

“…This is more accurate than the final joint training and classification through a single hidden layer based feed-forward neural network (FFNN)-for instance, a fully connected layer with either sigmoid activation or softmax classifier. To improve the ETD performance, we use Adaboost as a final classifier, which acts as a final layer of the LSTM-UNet to replace the single hidden layer based FFNN used in traditional models [20,22,23]. Therefore, Adaboost simply takes the outputs of LSTM-UNet modules and concatenates them to make a new input for the Adaboost model.…”

Section: Joint Training and Classification Modulementioning

confidence: 99%

“…In contrast to the manual methods, supervised machine learning solutions have gained the interest of utilities and academia for performing ETD. Studies based on the supervised learning techniques [20][21][22][23][24][25] focus on ETD using the large and imbalanced datasets obtained through the smart meters. However, the performances of these techniques are still not sufficient for the practical applications in utilities.…”

Section: Introductionmentioning

confidence: 99%

“…Li et al [40] introduce a hybrid of CNN and random forest (RF), known as the CNN-RF model, for ETD in smart grids, where the RF is used as the final layer to perform ETD on the extracted features. Hasan et al [22] made a solution consisting of the hybrid of CNN and LSTM for ETD. LSTM is used to solve the binary classification problem using CNN's output.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A Combined Deep Learning and Ensemble Learning Methodology to Avoid Electricity Theft in Smart Grids

et al. 2020

View full text Add to dashboard Cite

Electricity is widely used around 80% of the world. Electricity theft has dangerous effects on utilities in terms of power efficiency and costs billions of dollars per annum. The enhancement of the traditional grids gave rise to smart grids that enable one to resolve the dilemma of electricity theft detection (ETD) using an extensive amount of data formulated by smart meters. This data are used by power utilities to examine the consumption behaviors of consumers and to decide whether the consumer is an electricity thief or benign. However, the traditional data-driven methods for ETD have poor detection performances due to the high-dimensional imbalanced data and their limited ETD capability. In this paper, we present a new class balancing mechanism based on the interquartile minority oversampling technique and a combined ETD model to overcome the shortcomings of conventional approaches. The combined ETD model is composed of long short-term memory (LSTM), UNet and adaptive boosting (Adaboost), and termed LSTM–UNet–Adaboost. In this regard, LSTM–UNet–Adaboost combines the advantages of deep learning (LSTM-UNet) along with ensemble learning (Adaboost) for ETD. Moreover, the performance of the proposed LSTM–UNet–Adaboost scheme was simulated and evaluated over the real-time smart meter dataset given by the State Grid Corporation of China. The simulations were conducted using the most appropriate performance indicators, such as area under the curve, precision, recall and F1 measure. The proposed solution obtained the highest results as compared to the existing benchmark schemes in terms of selected performance measures. More specifically, it achieved the detection rate of 0.92, which was the highest among existing benchmark schemes, such as logistic regression, support vector machine and random under-sampling boosting technique. Therefore, the simulation outcomes validate that the proposed LSTM–UNet–Adaboost model surpasses other traditional methods in terms of ETD and is more acceptable for real-time practices.

show abstract

Section: Data Balancingmentioning

confidence: 99%

Section: Unet Modulementioning

confidence: 99%

Section: Joint Training and Classification Modulementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A Combined Deep Learning and Ensemble Learning Methodology to Avoid Electricity Theft in Smart Grids

et al. 2020

View full text Add to dashboard Cite

show abstract

Gradient boosting and Shapley additive explanations for fraud detection in electricity distribution grids

Santos¹,

Yamouni²,

Albiero³

et al. 2021

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

Fraud in electrical energy consumption represents a critical economic burden for utility companies around the world. Despite systematic efforts to mitigate electricity theft, this practice persists mostly in developing countries where companies rely on traditional detection methods. In Brazil it is estimated that around 7% of the total electrical energy available for consumption in 2016 was lost due to frauds. Here we describe an efficient and scalable system to predict fraudulent behavior and guide in loco inspections. We compared the performances of several machine learning algorithms using consumption and inspection data provided by CPFL Energia. We show that proper feature engineering and boosted classification trees trained with XGBoost are able to extract patterns related to fraud occurrence and to achieve predictive power of practical consequences. Moreover, we demonstrate how Shapley additive explanation (SHAP) values can be employed to build user friendly explanations. Together, the proposed model and its explainers contribute not only to reveal potentially fraudulent behavior but also to understand root causes, what can be used to devise robust mitigation strategies.

show abstract

Big data analytics for identifying electricity theft using machine learning approaches in microgrids for smart communities

Arif

Javaid

Aldegheishem

et al. 2021

Concurrency and Computation

View full text Add to dashboard Cite

Electricity theft (ET) causes major revenue loss in power utilities. It reduces the quality of supply, raises production cost, causes legal consumers to pay the higher cost, and impacts the economy as a whole. In this article, we use the State Grid Corporation of China (SGCC) dataset, which contains electricity consumption data of 1035 days for two classes: normal and fraudulent. In this work, ET detection model is proposed that consists of four steps: interpolation, data balancing, feature extraction, and classification. First, missing values of the dataset are recovered using the interpolation method. Second, resampling technique is implemented. ET consumers are 9% in the SGCC dataset that make the model inefficient to correctly classify both classes (normal and theft). A hybrid resampling technique is proposed, named synthetic minority oversampling technique with near miss. Third, residual network extracts the latent features from the SGCC dataset. Fourth, three tree based classifiers, such as decision tree (DT), random forest (RF), and adaptive boosting (AdaBoost) are applied to train the encoded feature vectors for classification. Besides, search for good hyperparameters is a challenging task, which is usually done manually and takes a considerable amount of time. To resolve this problem, Bayesian optimizer is used to simplify the tuning process of DT, RF, and AdaBoost. Finally, the results indicate that RF outperforms DT and AdaBoost.

show abstract

Electricity Theft Detection in Smart Grid Systems: A CNN-LSTM Based Approach

Cited by 283 publications

References 55 publications

A Combined Deep Learning and Ensemble Learning Methodology to Avoid Electricity Theft in Smart Grids

A Combined Deep Learning and Ensemble Learning Methodology to Avoid Electricity Theft in Smart Grids

Gradient boosting and Shapley additive explanations for fraud detection in electricity distribution grids

Big data analytics for identifying electricity theft using machine learning approaches in microgrids for smart communities

Contact Info

Product

Resources

About