A shape-based clustering method for pattern recognition of residential electricity consumption

Having investigated the determinants of electricity consumption in relation to household and dwelling characteristics, Chapter 6 provides a closer look at the behavioral patterns of household appliance use and electricity consumption. The OTB sample was used to conduct correlation and factor analysis. This Chapter deals with the Research Question III-2 of this thesis: (Chapter 1, Section 3, pg. 16-17) “ III. What are the behavioral patterns and profiles of energy consumption? The sub-question is: What are the behavioral patterns of electricity consumption? How do they relate to the household characteristics, revealing behavioral profiles? “ The research reported in this Chapter was conducted by Bedir. The data was collected by a questionnaire prepared by Guerra Santin and Bedir, using OTB’s means of data collection. The analysis was done, and the paper was written by Bedir. The co-author (E.C. Kara) commented on methodology of the research. The co-author has given his permission to include the paper in the thesis. This study was published in Energy and Buildings: Bedir, M. Kara, E.C. “Behavioral Patterns and Profiles of Electricity Consumption in Dutch Dwellings” Energy and Buildings, Available online 12 June 2017, http://dx.doi. org/10.1016/j.enbuild.2017.06.015

Background: Understanding customers’ electricity consumption patterns is essential for developing predictive analytics, which is needed for effective supply and demand management. Objective: This study aims to understand customers’ segmentation and consumption behaviour using a hybrid approach combining the K-Means clustering, customer lifetime value concept, and analytic hierarchy process. Methods: This study uses more than 16 million records of customers’ electricity consumption data from January 2019 to December 2020. The K-Means clustering identifies the initial market segments. The results were then evaluated and validated using the customer lifetime value concept and analytical hierarchy process. Results: Three customer segments were identified. Segment 1 has 282 business customers with a total capacity of 938,837 kWh, peak load usage of 27,827 kWh, and non-peak load usage of 115,194 kWh. Segment 2 has 508,615 business customers with a total capacity of 4,260 kWh, a peak load of 35 kWh, and a non-peak load of 544 kWh. Segment 3 has 37 business customers with a total capacity of 2,226,351 kWh, a peak load of 123.297 kWh, and a non-peak load of 390,803. Conclusion: A business strategy that could be taken is to base customer relationship management (CRM) on the three-customer segmentation. For the least profitable segment, aside from retail account marketing, a continuous partnership program is needed to increase electricity consumption during the non-peak period. For the highly and moderately profitable segments, a premium business-to-business approach can be applied to accommodate their increasing energy consumption without excessive electricity use in the peak period. Special account executives need to be deployed to handle these customers.

Behavioral patterns and profiles of electricity consumption in dutch dwellings

A new mining framework with piecewise symbolic spatial clustering

Data classification of users’ electricity consumption provides an in-depth analysis for users’ electricity consumption 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 electricity consumption data. In this paper, we propose a privacy-preserving classification scheme for electricity consumption data 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.

The literature does not present integral solutions to allow using the same credential to access the smart meter and smart house from an electric utility and vice-versa. The main reason being the technology gap in the communication between the Advanced Metering Infrastructure (AMI) and the Internet. The technology used in the Internet domain to communicate with Data Concentrators (DC) and the electric utility is more powerful than the technology used in the communication between smart meters and the DC, which is bandwidth limited and better suited to the Internet of Things (IoT) domain. Therefore, we are proposing the use of Identity Management (IdM) and a key-based scheme to enable the integration of IoT and the Internet using the same credentials, without creating a security bottleneck in the communication. An additional security mechanism is provided in the smart house context to isolate the house from direct accesses from the Internet, though allowing the utility to reconfigure the electric power consumption profile to avoid a potential blackout, for instance. Our proposal includes multi-sensor anti-tampering techniques to provide physical protection to a smart meter, in conjunction with a multilevel integrity mechanism to provide logical protection to its resource-constrained microcontroller, given the smart meter is a key component to mitigate electricity consumption fraud. The prototype has shown that our proposal is feasible for protecting the smart house, smart meter and the end-to-end communication between smart meter or house and the utility.

A data mining-based framework for the identification of daily electricity usage patterns and anomaly detection in building electricity consumption data

Extremely randomised trees machine learning model for electricity theft detection

As the global population ages, assisted living technologies for the elderly are becoming more popular. A person normally performs activities of daily living (ADLs) on a regular basis. A person's ability to perform recurring ADLs indicates the person's wellness. Anomalies in activity patterns of a person might indicate changes in the person's wellness. A method is proposed in this paper for detecting anomalies in activity patterns of a lone occupant using his/her electricity consumption data. The proposed method infers anomalies in activity patterns of an occupant from electricity consumption patterns instead of explicitly monitoring the underlying individual activities. The proposed method provides a score which is a quantitative assessment of anomalies in electricity consumption pattern of an occupant for a given day. A survey was conducted to obtain the hourly activities of three lone occupants for a month. From the survey, the level of suspicion values which are quantitative assessments of anomalies in activity patterns of the occupants were deduced. Using Fuzzy C-Means (FCM) clustering with Euclidean distance measure, the scores and level of suspicion values were clustered respectively. A day was then classified as regular or irregular from an electricity consumption perspective (score) and an activity perspective (level of suspicion value) respectively. Our results show that anomalies in electricity consumption patterns correlate well with anomalies in the underlying activity patterns.

In recent years, the problem of electrical load forecasting gained attention due to the arrival of new measurement technologies that produce electrical energy consumption data at very short intervals of time. Such short term measurements become voluminous in very short time. The availability of big electrical consumption data allows machine learning techniques to be employed to analyze consumption behavior of every consumer on a greater detail. Predicting the consumption of a residential customer is crucial at this point of time because tailor-made consumer-specific tariffs will play a vital role in load balancing process of Utilities. This paper analyzes the electrical consumption of a single residential customer measured using a smart meter that is capable of measuring electrical consumption at circuit level. The issues and challenges in collecting the data and pre-processing required for making them suitable for data analytics are discussed in detail. A comparison of the performance of different machine learning algorithms implemented using Python’s Scikit-learn module gives an insight on the consumption pattern.

With the development of IoT technology, smart grid has gradually replaced the traditional grid. Smart grid is convenient and fast. It can provide real-time residential electricity monitoring and forecasting, give users better electricity guidance and save a lot of labor costs. Smart meters send customers’ electricity consumption data to the gateway, which aggregates the data and then sends it to the electricity consumption control center. But in this process, there will be a security problem of leakage of customer’s electricity consumption data. Most of the current user data privacy protection collection schemes use homomorphic encryption and randomization techniques. However, some of these schemes require a trusted third-party entity, and some may cause significant computational overhead. Due to the limited computational resources of smart meters, these techniques may be impractical. In this paper, we propose a local differential privacy data aggregation protection scheme based on the idea of grouping perturbation of electricity consumption data according to data domains. Experiments show that our scheme can provide statistical estimates of electricity in the region while satisfying the privacy protection of customers’ electricity consumption data. Moreover, our scheme has small computational and communication overheads, which can meet the application requirements in practical scenarios.

Identifying household electricity consumption patterns: A case study of Kunshan, China

In Smart Grids the Information and Communication Technologies (ICT) could be used to better manage both consumption and production of electricity. The increasing presence of renewable energy sources in production and the permeation of novel consumption types (e.g. Plug-in Hybrid Electric Vehicles (PHEV)) will obviously cause the increase the fluctuation of electrical energy. One possible solution to these problems is development of novel methods for investigating electrical power consumption data series. As the existing learning algorithms of pattern classification are suitable for discovering internal structures of large datasets, it is important to generate a training/testing/validation learning database from existing measurements (e.g. from smart meters), actually via segmentation and labeling by hand. In this paper we propose a novel method for the automatic segmentation with a predefined confidence level. The algorithm is based on the generalized Jensen-Shannon divergence (JSD), and it estimates the change-points (CPTs) in electrical power consumption data. Both the method and some recent results in segmenting one household's power consumption data are presented in this paper.

Relevance. The increasing electricity consumption in the private residential sector, driven in part by the growing use of electric heating, is leading to higher loads on 0.4 kV power transmission lines. Traditional standardized load profiles do not always reflect modern consumption patterns and conditions, which creates risks of inaccurate assessments of the electrical grid’s capacity and necessitates more precise modeling of grid operating conditions.Purpose. To develop approaches for classifying consumers and identifying statistically significant patterns in electricity consumption in private residential areas for subsequent calculation of grid operating conditions.Methods. The analysis was based on half-hourly electricity consumption data from 42 private houses, collected via an Automated Meter Reading and Management System (AMRMS). The data was cleaned using the three-sigma rule to remove gaps and outliers, and heat maps were used to identify non-representative consumers. The statistical significance of differences was determined using analysis of variance (ANOVA) and Tukey’s test. Based on median consumption values, consumer groups were formed (low and high electricity consumption). Data processing and visualization were performed using MS Excel, Python (Pandas, NumPy, SciPy libraries), and the Statistica software package.Results. The analysis confirmed statistically significant differences in electricity consumption between most of the houses (F = 2065.4, p < 0.001). Tukey’s test showed that within each group, homes exhibited relatively stable energy consumption values, while intergroup comparisons revealed substantial variations in electricity usage. As a result of the study, two consumer types were identified: "low" and "high" consumption groups. The high-consumption group exhibited distinct evening peaks (18:00–22:00), whereas the low-consumption group had a more evenly distributed load profile.Conclusion. The application of statistical analysis methods to electricity consumption data enabled the simplification of household classification into two main groups and the development of typical consumption profiles. These results were integrated into the LineCapacity software, facilitating grid operation calculations and reducing the risk of misjudging the available power transmission capacity. A promising research direction is planned, focusing on expanding the dataset on residential electricity consumption. This will allow for the consideration of seasonal factors and the development of simulation modeling mechanisms for various consumer groups.

Understanding activities within premises can contribute to a fuller understanding of energy use within buildings and the building stock. Analysis of detailed surveys of over 300 non-domestic premises has produced empirical room-scale space-use profiles for 16 premises types. Electricity consumption and internal gains resulting from the operation of electrical appliances have been characterized for 295 combinations of internal space use and premises activity type. For each combination, the outputs include the energy-use intensity (kWh/m2/year) profiles for: overall appliance consumption; 14 end uses of appliances (e.g. lighting, catering, computing); and 18 groups of appliance activity descriptions (e.g. sales, office work, process). These profiles of characteristics were created for application in an urban-scale energy-use model, based upon premises floor space recorded in property taxation data, without the need for detailed energy surveys of premises. Appliance electricity consumption and internal gains are revealed at a finer spatial resolution than previous methods, indicating the diversity of energy-use characteristics in greater detail than for entire homogenized premises or premises types. This method may be used for evaluating the current electricity consumption and consequent carbon emissions from the non-domestic building stock (or parts thereof) and for estimating the effects of potential interventions.