Electricity Consumption Data Research Articles

A hybrid model for deep learning short-term power load forecasting based on feature extraction statistics techniques

Accurate and reliable load forecasting can ensure the safety and economy of power system operation. To improve the accuracy of short-term power load forecasting, this paper adopts feature extraction combined with a variety of machine learning methods,(Empirical wavelet decomposition (EWT), Convolutional neural network (CNN) model, Recurrent neural network (RNN) model, Long short-term memory network (LSTM) model, Bayesian optimization (BOA) algorithm). A new hybrid forecasting model, the EWT-CNN-S-RNN + LSTM model is proposed to predict power consumption. In this method, the fixed mode of EWT decomposition is used to extract statistical features, and the LSTM/RNN model is selected according to the statistical features. This method uses the Bayesian optimization (BOA) algorithm to optimize the parameters to solve the model gradient explosion problem. Based on the electricity consumption data of 2007 (Australian Energy Market Operator) and a total of 336 power load data, a good load forecasting result is obtained, which can provide indirect support for power regulation and sustainable development of electricity.

Novel hybrid chaotic map-based secure data transmission between smart meter and HAN devices

ABSTRACT Home area network (HAN) devices send the electricity consumption and other important data to the smart meter that must remain confidential from other devices. This paper proposes a novel one-dimensional hybrid chaotic map. The proposed map shows excellent chaotic properties when analyzed by bifurcation diagram, Lyapunov exponent & Shannon entropy. We further design an encryption strategy for data transfers between the smart meter and HAN devices. The proposed encryption scheme uses the existing lightweight key management in advanced metering infrastructure (LKM-AMI) architecture for data transfers, in which the encrypted data is transferred through an insecure channel and private keys are provided by trusted third party (TTP) through secure channels. The 2-way communication between HAN devices and the smart meter sends messages that are encrypted by using the proposed novel hybrid one-dimensional chaotic map. The encryption strategy mainly consists of three steps. In the first step, the seed and the control parameters are initialized. The second phase generates two intermediate keys using the proposed hybrid chaotic map. In the last phase, we encrypt the message by applying permutation followed by diffusion using intermediate keys. The proposed encryption strategy is resistant to various attacks.

Privacy-Preserving Electricity Data Classification Scheme Based on CNN Model With Fully Homomorphism

Data classification of users' electricity consump- tion provides an in-depth analysis for users' electricity con- sumption status, which plays a vital role in the management and distribution of electric energy. So, some data classification methods have been proposed to solve the classification problem of electricity consumption data. However, plaintext-based data classification may bring about the privacy leakage of electric- ity consumption data. In this paper, we propose a privacy- preserving classification scheme for electricity consumption da- ta under fog computing-based smart metering system, which is based on convolutional neural network (CNN) model with fully homomorphic method (CKKS). The target of our proposed scheme is to solve the leakage problem of private electricity consumption data during the classification procedure. In our scheme, an improved K-means-based labeling algorithm is constructed to process historical electricity consumption data, which is used as the sample data to train the CNN classification model by cloud server. Also, the fog nodes are only permitted to obtain the related ciphertext parameters of the trained CNN model, and perform the classification of ciphertext-based electricity consumption data generated by fully homomorphic method. Based on the classical testing data, the experimental results show that our proposed classification scheme can provide the high classification accuracy of electricity data while protecting the privacy of electricity data.

Anti-interference Technology of Multi-type Data Acquisition and Transmission Based on HPLC

With the development of the social economy, electricity consumption increases rapidly. A large amount of real-time electricity consumption data plays an important role in analyzing the power grid business. Monitoring power consumption performance through electricity consumption data is also an important indicator to improve power grid performance. By analyzing the characteristics of HPLC channels, such as capacitance characteristics, frequency loss characteristics, signal-to-noise ratio characteristics, line icing, and noise interference, the critical values of data transmission distortion caused by the characteristics of each channel were summarized. On this basis, it is concluded that suitable filters and amplifiers should be selected to improve the signal-to-noise ratio and the transmission quality of HPLC signals. Therefore, an anti-interference technology for data transmission based on HPLC is proposed. Finally, the application advantages and existing problems of HPLC technology in three-stream integration data acquisition and transmission are further explored.

Does high speed railway alleviate housing vacancy rates? Evidence from smart meter data of household electricity consumption

Does high speed railway alleviate housing vacancy rates? Evidence from smart meter data of household electricity consumption

Step towards secure and reliable smart grids in Industry 5.0: A federated learning assisted hybrid deep learning model for electricity theft detection using smart meters

The integration of Smart Grid technology and conceptual Industry 5.0 has paved the way for advanced energy management systems that enhance efficiency and revolutionized the parallel integration of power sources in a sustainable manner. However, this digitization has opened a new stream of the threat and opportunities of electricity theft posing a significant challenge to the security and reliability of Smart Grid networks. In this paper, we propose a secure and reliable theft detection technique using deep federated learning (FL) mechanism. The technique leverages the collaborative power of FL to train a Convolutional Gated Recurrent Unit (ConvGRU) model on distributed data sources without compromising data privacy. The training deep learning model backbone consists of a ConvGRU model that combines convolutional and gated recurrent units to capture spatial and temporal patterns in electricity consumption data. An improvised preprocessing mechanism and hyperparameter tuning is done to facilitate FL mechanism. The halving randomized search algorithm is used for hyperparameters tuning of the ConvGRU model. The impact of hyperparameters involved in the ConvGRU model such as number of layers, filters, kernel size, activation function, pooling, GRU layers, hidden state dimension, learning rate, and the dropout rate is elaborated. The proposed technique achieves promising results, with high accuracy, precision, recall, and F1 score, demonstrating its efficacy in detecting electricity theft in Smart Grid networks. Comparative analysis with existing techniques reveal the superior performance of the deep FL-based ConvGRU model. The findings highlight the potential of this approach in enhancing the security and efficiency of Smart Grid systems while preserving data privacy.

Identifying rural high energy intensity residential buildings using metered data

The residential housing stock in rural areas of the U.S. is often older and less energy-efficient than homes built in urban areas. To characterize the energy performance of these buildings and facilitate their energy efficiency upgrades, this study provides a framework for analyzing building energy time-series data to identify buildings with similar temporal energy performance patterns. Using an unsupervised hierarchical time-series clustering algorithm, this framework aims to identify highly inefficient buildings as strong candidates for energy efficiency improvement programs. Weather-normalized electricity and natural gas consumption data from 320 properties in the rural town of Bloomfield, Iowa, covering a six-year reporting period from 2014 to 2019 is used to optimally cluster energy consumption data into two groups, including average and high intensity energy users. To correct the underlying bias of the commonly used area-normalized energy use intensity (EUI) indexes towards smaller homes, the proposed clustering algorithm only compares the EUI of homes with similar floor areas. User types are defined for electricity and natural gas consumption separately to allow for tailored actions and policy recommendations. Accordingly, 5% and 7% of residences in the dataset were classified as high intensity users for electricity and natural gas, respectively. Of the remaining, 88% of homes were classified as average intensity users for both energy types. Only 2 properties in the dataset (less than 1%) were identified to be high intensity users for both electricity and natural gas. Homes classified as high intensity electricity and natural gas users, although smaller in number, were, on average, 44% and 30% more energy intensive, respectively, when compared to similar-sized average intensity consumers. The results also indicated that older homes are generally more likely to be high intensity natural gas consumers while a relatively higher percentage of newer homes were classified as high intensity electricity users.

Analyzing Long-Term and High Instantaneous Power Consumption of Buildings from Smart Meter Big Data with Deep Learning and Knowledge Graph Techniques

In the context of the growing emphasis on energy conservation and carbon reduction, the widespread deployment of smart meters in residential and commercial buildings is instrumental in promoting electricity savings. In Taiwan, local governments are actively promoting the installation of smart meters, empowering residents to monitor their electricity consumption and detect abnormal usage patterns, thus mitigating the risk of electrical fires. This safety-oriented approach is a significant driver behind the adoption of smart meters. However, the analysis of the substantial data generated by these meters necessitates pre-processing to address anomalies. Presently, these data primarily serve billing calculations or the extraction of power-saving patterns through big data analytics. To address these challenges, this study proposes a comprehensive approach that integrates a relational database for storing electricity consumption data with knowledge graphs. This integrated method effectively addresses data scarcity at various time scales and identifies prolonged periods of excessive electricity consumption, enabling timely alerts to residents for specific appliance shutdowns. Deep learning techniques are employed to analyze historical consumption data and real-time smart meter readings, with the goal of identifying and mitigating hazardous usage behavior, consequently reducing the risk of electrical fires. The research includes numerical values and text-based predictions for a comprehensive evaluation, utilizing data from ten Taiwanese households in 2022. The anticipated outcome is an improvement in household electrical safety and enhanced energy efficiency.

Deep learning-based meta-learner strategy for electricity theft detection

Electricity theft damages power grid infrastructure and is also responsible for huge revenue losses for electric utilities. Integrating smart meters in traditional power grids enables real-time monitoring and collection of consumers’ electricity consumption (EC) data. Based on the collected data, it is possible to identify the normal and malicious behavior of consumers by analyzing the data using machine learning (ML) and deep learning methods. This paper proposes a deep learning-based meta-learner model to distinguish between normal and malicious patterns in EC data. The proposed model consists of two stages. In Fold-0, the ML classifiers extract diverse knowledge and learns based on EC data. In Fold-1, a multilayer perceptron is used as a meta-learner, which takes the prediction results of Fold-0 classifiers as input, automatically learns non-linear relationships among them, and extracts hidden complicated features to classify normal and malicious behaviors. Therefore, the proposed model controls the overfitting problem and achieves high accuracy. Moreover, extensive experiments are conducted to compare its performance with boosting, bagging, standalone conventional ML classifiers, and baseline models published in top-tier outlets. The proposed model is evaluated using a real EC dataset, which is provided by the Energy Informatics Group in Pakistan. The model achieves 0.910 ROC-AUC and 0.988 PR-AUC values on the test dataset, which are higher than those of the compared models.

Polycentric cities and income polarization in China: Does city spatial structure matter?

Income polarization is a critical issue of social development. This study analyzes the role of polycentric cities in alleviating income polarization using monthly household electricity consumption data to identify income polarization based on the application of proxy variable analysis and an artificial neural network. The polycentric structure of cities is identified using unique satellite nighttime light data. It is found that polycentric structure is significant for alleviating income polarization, which exhibits a nonlinear effect. As the polycentric structure index increases, the inhibitory effect of polycentric cities on income polarization will increase. Relevant policy implications are proposed based on the study's findings.

EWELD: A Large-Scale Industrial and Commercial Load Dataset in Extreme Weather Events

Load forecasting is crucial for the economic and secure operation of power systems. Extreme weather events, such as extreme heat and typhoons, can lead to more significant fluctuations in power consumption, making load forecasting more difficult. At present, due to the lack of relevant public data, the research on load forecasting under extreme weather events is still blank, so it is necessary to release a large-scale load dataset containing extreme weather events. The dataset includes electricity consumption data of industrial and commercial users under extreme weather events such as typhoons and extreme heat, which are collected at 15-minute intervals. The data is collected over six years from smart meters installed at the power entry points of users in southern China. The dataset consists of electricity consumption data from 386 industrial and commercial users in 17 industries, with more than 50 million records. During the recording period, extreme weather events such as typhoons and extreme heat are marked to form a total of 5,741 event records.

Benford’s Law in Electric Distribution Network

Benford’s law can be used as a method to detect non-natural changes in data sets with certain properties; in our case, the dataset was collected from electricity metering devices. In this paper, we present a theoretical background behind this law. We applied Benford’s law first digit probability distribution test for electricity metering data sets acquired from smart electricity meters, i.e., the natural data of electricity consumption acquired during a specific time interval. We present the results of Benford’s law distribution for an original measured dataset with no artificial intervention and a set of results for different kinds of affected datasets created by simulated artificial intervention. Comparing these two dataset types with each other and with the theoretical probability distribution provided us the proof that with this kind of data, Benford’s law can be applied and that it can extract the dataset’s artificial manipulation markers. As presented in the results part of the article, non-affected datasets mostly have a deviation from BL theoretical probability values below 10%, rarely between 10% and 20%. On the other side, simulated affected datasets show deviations mostly above 20%, often approximately 70%, but rarely lower than 20%, and this only in the case of affecting a small part of the original dataset (10%), which represents only a small magnitude of intervention.

Quantifying Photovoltaic surplus at an urban scale: A case study in Seoul

With the increasing photovoltaic (PV) penetration in urban buildings, the temporal overproduction of PV electricity presents both a challenge and an opportunity for solar energy utilization in high-PV-penetration urban environments. While existing research has focused on strategies for improving self-consumption, little to no information exists on the magnitude and implications of PV surplus at the urban level. Questions such as when and where PV surplus occurs in the urban area remain unanswered, and there is a lack of established knowledge about how urban contexts relate to PV surplus. Therefore, this study aimed to quantify the PV surplus in a high roof-PV penetration scenario at an urban scale, using a method that couples hourly PV production and electricity consumption data from individual buildings in urban areas. The paper explored two aspects of PV surplus, namely PV surplus hours (PVsH) and PV surplus energy (PVsE), which were utilized to identify the temporal occurrence and quantify the amount of PV surplus, respectively. One residential and one office area in Seoul were examined. Results showed abundant PV surplus resources in case areas, especially in residential buildings. A wide range of surplus distribution was found among the buildings, especially in the office area. The PV surplus peaked between late spring and early summer. A noticeable correlation between building morphology and PVsE was detected. The results suggested the PV surplus of the residential area is sufficient for localizing charging demand by comparing PV surplus against the electricity demand from a hypothetical scenario of a full electric vehicle (EV) community.

Electric Vehicle Trip Information Inference Based on Time-Series Residential Electricity Consumption

Pure electric vehicles (EVs) become more and more popular in current automotive markets. With wide application of EVs in the market, more EVs will be charged in residential homes. An advanced metering infrastructure measures real-time electricity consumption of individual residential homes. However, the advanced metering infrastructure is exposed to attacks given the fact that its components (e.g., smart meters) are located at public places. Under this situation, electricity consumption information of an individual residential home with an EV can be easily stolen. In this paper, we propose an EV Trip Information Inference System (TIIS) to infer EV trip information (origin-destination of each driving trip in a day) of a residential home based on its real-time electricity consumption measurements in a day. To the best of our knowledge, this is the first driving privacy relevant work caused only by residential electricity consumption data. For a residential home with an EV, TIIS derives the relationship between trip numbers in a day and charging-starting time, and the relationship between total driving distance and total charging energy. To derive the EV trip information of a residential home, TIIS firstly extracts an EV charging profile from its electricity consumption data and identifies EV model type and trip numbers as well as total driving distance based on EV charging profiles. Then, TIIS infers EV trip information based on the identified information. We used electricity consumption data and EV driving data to evaluate inference performance of TIIS. The experimental results demonstrate TIIS has as high as 81% trip inference accuracy and works on residential homes with different EV model types.

Machine learning-based estimation of buildings' characteristics employing electrical and chilled water consumption data: Pipeline optimization

Machine learning-based estimation of buildings' characteristics employing electrical and chilled water consumption data: Pipeline optimization

Assessment of the impact of COVID-19 lockdown measures on electricity consumption – Evidence from Portugal and Spain

The COVID-19 pandemic caused unheard-of shifts in socioeconomic norms and interpersonal connections. Governments were compelled to place limitations on the way of life of residents and companies in order to stem the virus' spread in order to deal with this new threat. These led to stay-at-home orders in many nations and reduced industry activities to their bare minimum, which had an immediate impact on the electricity systems.This article investigates the influence of COVID-19 on electricity consumption in Portugal and Spain, with the goal of shedding light on consumption shifts and how these are related to the stringency degree of government-imposed measures, as well as providing an overview of the first post-pandemic data. We cross information between observed electricity consumption data and numerous factors, including the Stringency Index (SI) given by the Oxford COVID-19 Government Response Tracker (OxCGRT), using an econometric model. We also investigate how government-issued alert levels influenced electricity use.We concluded that, although Spain only felt the repercussions in 2020, Portugal saw a negative and considerable impact from restriction measures and lockdowns in 2020 and 2021. The decision to close schools was found to be the specific step that significantly reduced electricity consumption in both countries.

The Use of Real Energy Consumption Data in Characterising Residential Energy Demand with an Inventory of UK Datasets

The availability of empirical energy data from Advanced Metering Infrastructure (AMI)—which includes household smart meters—has enabled residential energy demand to be characterised in different forms. This paper first presents a literature review of applications of measured electricity, gas, and heat consumption data at a range of temporal resolutions, which have been used to characterise and develop an understanding of residential energy demand. User groups, sectors, and policy areas that can benefit from the research are identified. Multiple residential energy demand datasets have been collected in the UK that enable this characterisation. This paper has identified twenty-three UK datasets that are accessible for use by researchers, either through open access or defined processes, and presents them in an inventory containing details about the energy data type, temporal and spatial resolution, and presence of contextual physical and socio-demographic information. Thirteen applications of data relating to characterising residential energy demand have been outlined in the literature review, and the suitability of each of the twenty-three datasets was mapped to the thirteen applications. It is found that many datasets contain complementary contextual data that broaden their usefulness and that multiple datasets are suitable for several applications beyond their original project objectives, adding value to the original data collection.

Towards Efficient Energy Consumption: Audit of Households Electricity Consumption in Ghana

Electricity from national grids remain the most reliable and affordable source of power to many households globally. However, in Ghana, the supply of electricity from the national grid to households is generally unstable with residential electricity tariff skyrocketing. Consumers are unable to quantify the amount of electricity purchased in kwh/month in order to appreciate the tariff band they belong to. The billing authorities only account for the approximate unit (kwh/month) for the nearest cedi spent on electricity. The run-offs lead to over/under-billing of consumers for electricity purchased and, thus, derail the agenda of efficient energy management. Given this background, we proceeded to investigate households’ electricity consumption by first, estimating the exact consumption units of electricity for any amount of electricity power purchased. Subsequently, we explored factors that affected households’ electricity consumption in Ghana. We relied on 8,599 households’ data captured in the Ghana Living Standard Survey 7. The method of maximum likelihood estimation was used to estimate a multiple regression model fitted to the monthly electricity consumption data. The results showed that the fewer are alternative sources of electricity available to households, the greater is the consumption of electricity by the households. Other significant variables found were number of rooms in a household, area (m2) of the house, number of people in the house, categories of tariff bands and number of electrical appliances in use. These variables showed positive relationship with the natural log of the estimated electricity consumed per month for any household.

A Power Load Forecasting Method Based on Intelligent Data Analysis

Abnormal electricity consumption behavior not only affects the safety of power supply but also damages the infrastructure of the power system, posing a threat to the secure and stable operation of the grid. Predicting future electricity consumption plays a crucial role in resource management in the energy sector. Analyzing historical electricity consumption data is essential for improving the energy service capabilities of end-users. To forecast user energy consumption, this paper proposes a method that combines adaptive noise-assisted complete ensemble empirical mode decomposition (CEEMDAN) with long short-term memory (LSTM) networks. Firstly, considering the challenge of directly applying prediction models to non-stationary and nonlinear user electricity consumption data, the adaptive noise-assisted complete ensemble empirical mode decomposition algorithm is used to decompose the signal into trend components, periodic components, and random components. Then, based on the CEEMDAN decomposition, an LSTM prediction sub-model is constructed to forecast the overall electricity consumption by using an overlaying approach. Finally, through multiple comparative experiments, the effectiveness of the CEEMDAN-LSTM method is demonstrated, showing its ability to explore hidden temporal relationships and achieve smaller prediction errors.

Electricity theft detection for energy optimization using deep learning models

AbstractThe rapid increase in nontechnical loss (NTL) has become a principal concern for distribution system operators (DSOs) over the years. Electricity theft makes up a major part of NTL. It causes losses for the DSOs and also deteriorates the quality of electricity. The introduction of advanced metering infrastructure along with the upgradation of the traditional grids to the smart grids (SGs) has helped the electric utilities to collect the electricity consumption (EC) readings of consumers, which further empowers the machine learning (ML) algorithms to be exploited for efficient electricity theft detection (ETD). However, there are still some shortcomings, such as class imbalance, curse of dimensionality, and bypassing the automated tuning of hyperparameters in the existing ML‐based theft classification schemes that limit their performances. Therefore, it is essential to develop a novel approach to deal with these problems and efficiently detect electricity theft in SGs. Using the salp swarm algorithm (SSA), gate convolutional autoencoder (GCAE), and cost‐sensitive learning and long short‐term memory (CSLSTM), an effective ETD model named SSA–GCAE–CSLSTM is proposed in this work. Furthermore, a hybrid GCAE model is developed via the combination of gated recurrent unit and convolutional autoencoder. The proposed model comprises five submodules: (1) data preparation, (2) data balancing, (3) dimensionality reduction, (4) hyperparameters' optimization, and (5) electricity theft classification. The real‐time EC data provided by the state grid corporation of China are used for performance evaluations via extensive simulations. The proposed model is compared with two basic models, CSLSTM and GCAE–CSLSTM, along with seven benchmarks, support vector machine, decision tree, extra trees, random forest, adaptive boosting, extreme gradient boosting, and convolutional neural network. The results exhibit that SSA–GCAE–CSLSTM yields 99.45% precision, 95.93% F1 score, 92.25% accuracy, and 71.13% area under the receiver operating characteristic curve score, and surpasses the other models in terms of ETD.