Energy Consumption Of Appliances Research Articles

A Nonintrusive Load Monitoring Method for Microgrid EMS Using Bi-LSTM Algorithm

Nonintrusive load monitoring in smart microgrids aims to obtain the energy consumption of individual appliances from the aggregated energy data, which is generally confronted with the error identification of the load type for energy disaggregation in microgrid energy management system (EMS). This paper proposes a classification strategy for the nonintrusive load identification scheme based on the bilateral long-term and short-term memory network (Bi-LSTM) algorithm. The sliding window algorithm is used to extract the detected load event features and obtain the load features of data samples. In order to accurately identify these load features, the steady state information is combined as the input of the Bi-LSTM model during training. Comprising long-term and short-term memory (LSTM) network and recurrent neural network (RNN), Bi-LSTM has the advantages of stronger recognition ability. Finally, precision (P), recall (R), accuracy (A), and F1 values are used as the evaluation method for nonintrusive load identification. The experimental results show the accuracy of the Bi-LSTM identification method for load start and stop state feature matching; moreover, the method can identify relatively low-power and multistate appliances.

PUMPNET: a deep learning approach to pump operation detection

Non-urgent high energy-consuming residential appliances, such as pool pumps, may significantly affect the peak to average ratio (PAR) of energy demand in smart grids. Effective load monitoring is an important step to provide efficient demand response (DR) to PAR. In this paper, we focus on pool pump analytics and present a deep learning framework, PUMPNET, to identify the pool pump operation patterns from power consumption data. Different from conventional time-series based Non-intrusive Load Monitoring (NILM) methods, our approach transfers the time-series data into image-like (date-time matrix) data. Then a U-shaped fully convolutional neural network is developed to detect and segment the image-like data in pixel level for operation detection. Our approach identify whether pool pumps operate given thirty-minute interval aggregated active power consumption data in kilowatt-hours only. Furthermore, the PUMPNET algorithm could identify pool pump operation status with high accuracy in the low-frequency sampling scenario for thousands of household, compared to traditional NILM algorithms which process high sampling rate data and can only apply to limited number of households. Experiments on real-world data validate the promising results of the proposed PUMPNET model.

An Improved Non-Schedulable Load Forecasting Strategy for Enhancing the Performance of the Energy Management in a Nearly Zero Energy Building

This paper proposes an improved non-schedulable load forecasting (NLF) technique that can be utilized to enhance the performance of the energy management system (EMS) in a nearly zero energy building (nZEB). The suggested NLF is based on a long short-term memory (LSTM) framework in conjunction with a semi-supervised clustering (SSC) technique, considering the most important features which may affect the energy consumption of the non-schedulable appliances (NAs), i.e. number and identity of residents, energy consumption, weather, temperature, humidity, outdoor irradiance, correlation with other loads, day of the week and holidays. The SSC algorithm is utilized to fill the uncompleted information for the residents’ presence in the house and its output constitutes one of the inputs of the LSTM based technique which provides as output a set of forecasting sequences of the NAs’ energy consumption. Unlike the published techniques, the proposed NLF method is not only based on the modeling of the residents’ preferences and habits, but it considers them as variables which affect the nZEB’s microgrid and EMS performance. Therefore, it predicts the residents’ behavior considering its interdependence with the nZEB’s microgrid, which can considerably contribute to the enhancement of the EMS effectiveness and performance. For the implementation of the proposed NLF, no additional hardware is required, but only amendments in the EMS to consider the NLF’s outcomes. To validate the effectiveness of the proposed NLF, selective Hardware-in-the-Loop results from a real nZEB are presented.

Swarm and evolutionary algorithms for energy disaggregation: challenges and prospects

Energy disaggregation is defined as the process of estimating the individual electrical appliance energy consumption of a set of appliances in a house from the aggregated measurements taken at a single point or limited points. The energy disaggregation problem can be modelled both as pattern recognition problem and as an optimisation problem. Among the two, the pattern recognition problem has been considerably explored while the optimisation problem has not been explored to the potential. In literature, researchers have attempted to solve the problem using various optimisation algorithms including swarm and evolutionary algorithms. However, the focus on optimisation-based methodologies, in general, swarm and evolutionary algorithm-based methodologies in particular is minimal. By considering the different problem formulations in the literature, we propose a framework to solve the energy disaggregation problem with swarm and evolutionary algorithms. With the help of simulation results using the existing problem formulations, we discuss the challenges posed by the energy disaggregation to swarm and evolutionary algorithm-based methodologies and analyse the prospects of these algorithms for the problem of energy disaggregation with some future directions.

Swarm and evolutionary algorithms for energy disaggregation: challenges and prospects

Energy disaggregation is defined as the process of estimating the individual electrical appliance energy consumption of a set of appliances in a house from the aggregated measurements taken at a single point or limited points. The energy disaggregation problem can be modelled both as pattern recognition problem and as an optimisation problem. Among the two, the pattern recognition problem has been considerably explored while the optimisation problem has not been explored to the potential. In literature, researchers have attempted to solve the problem using various optimisation algorithms including swarm and evolutionary algorithms. However, the focus on optimisation-based methodologies, in general, swarm and evolutionary algorithm-based methodologies in particular is minimal. By considering the different problem formulations in the literature, we propose a framework to solve the energy disaggregation problem with swarm and evolutionary algorithms. With the help of simulation results using the existing problem formulations, we discuss the challenges posed by the energy disaggregation to swarm and evolutionary algorithm-based methodologies and analyse the prospects of these algorithms for the problem of energy disaggregation with some future directions.

EnerGAN++: A Generative Adversarial Gated Recurrent Network for Robust Energy Disaggregation

Energy disaggregation, namely the separation of the aggregated household energy consumption signal into its additive sub-components, bears resemblance to the signal (source) separation problem and poses several challenges, not only as an ill-posed problem, but also, due to unsteady appliance signatures, abnormal behaviour that is usually detected in appliances operation and the existence of noise in the aggregated signal. In this paper, we propose EnerGAN++, a model based on Generative Adversarial Networks (GAN) for robust energy disaggregation. We attempt to unify the autoencoder (AE) and GAN architectures into a single framework, in which the autoencoder achieves a non-linear power signal source separation. EnerGAN++ is trained adversarially using a novel discriminator, to enhance robustness to noise. The discriminator performs sequence classification, using a recurrent convolutional neural network to handle the temporal dynamics of an appliance energy consumption time series. In particular, the proposed architecture of the discriminator leverages the ability of Convolutional Neural Networks (CNN) in rapid processing and optimal feature extraction, among with the need to infer the data temporal character and time dependence. Experimental results indicate the proposed method's superiority compared to the current state of the art.

Appliance-based solar certificates: A new way of going green

In recent years, an increasing amount of solar energy has been produced around the globe, but too few consumers actually buy it. Three studies now show how this purchase can be promoted with an approach similar to the carbon offset program in aviation. After a fictitious purchase of differently priced appliances (e.g. electric toothbrush or TV set) in an online store, participants could buy a solar certificate for this appliance, whereby its price depended on the appliance's energy consumption as well as the selected term (e.g. 1 or 3 years) of the certificate. By purchasing such a certificate, the amount of solar energy consumed by this appliance for the selected term would be fed into the Swiss electricity grid. Study 1 showed that participants were willing to buy such a certificate, especially certificates with longer terms. That willingness, however, depended on the “solar certificate price/appliance price” ratio. While Study 2 showed that this purchase is influenced by promotions (e.g., a specific reduction in price), Study 3 showed that this purchase did not lead to a less environmentally friendly behavior afterwards. Appliance-based solar certificates easily available at the point of sale can encourage consumers to buy more solar energy.

Federated Reinforcement Learning for Energy Management of Multiple Smart Homes With Distributed Energy Resources

This article proposesa novel federated reinforcement learning (FRL) approach for the energy management of multiple smart homes with home appliances, a solar photovoltaic system, and an energy storage system. The novelty of the proposed FRL approach lies in the development of a distributed deep reinforcement learning (DRL) model that consists of local home energy management systems (LHEMSs) and a global server (GS). Using energy consumption data, DRL agents for LHEMSs construct and upload their local models to the GS. Then, the GS aggregates the local models to update a global model for LHEMSs and broadcasts it to the DRL agents. Finally, the DRL agents replace the previous local models with the global model and iteratively reconstruct their local models. Simulation results obtained under heterogeneous home environments indicate the advantage of the proposed approach in terms of convergence speed, appliance energy consumption, and number of agents.

Internet of things and cloud computing‐based energy management system for demand side management in smart grid

A smart grid is an electricity network, which deals with electronic power conditioning and control of production, transmission, and distribution of electrical power by employing digital communication technologies to monitor and manage local changes in electricity usage. In the traditional power grid, energy consumers remain oblivious to their power consumption patterns, resulting in wasted energy as well as money. This issue is severely pronounced in the developing countries where there is a huge gap between demand and supply, resulting in frequent power outages and load-shedding. For electrical energy savings, the smart grid employs demand side management (DSM), which refers to adaptation in consumer's demand for energy through various approaches such as financial incentives and awareness. The DSM in future smart grid must exploit automated energy management systems (EMS) built upon the state-of-the-art technologies such as the internet of things (IoT) and cloud and/or fog computing. In this paper, we present the architecture framework, design, and implementation of an IoT and cloud computing-based EMS, which generates load profile of consumer to be accessed remotely by utility company or by the consumer. The consumers' load profiles enable utility companies to regulate and disseminate their incentives and incite the consumers to adapt their energy consumption. Our designed EMS is implemented on a Project Circuit Board (PCB) to be easily installed at the consumer premises where it performs the following tasks: (a) monitors energy consumption of electrical appliances by means of our designed current and voltage sensors, (b) uploads sensed data to Google Firebase cloud over many-to-many IoT communication protocol Message Queuing Telemetry Transport (MQTT) where consumer's load profile is generated, which can be accessed via a web portal. These load profiles serve as input for implementing the various DSM approaches. Our results demonstrate generated load profiles of consumer load in terms of current, voltage, energy, and power accessible via a web portal.

Experimental Evaluation and Kinetic Analysis of Direct-Contact Ultrasonic Fabric Drying Process

Abstract Over the past two decades, due to the rising energy prices and growing awareness about climate change, significant efforts have been devoted to reducing the energy consumption of various home appliances. However, the energy efficiency of clothes dryers has little improvement. Recent innovations in the direct-contact ultrasonic fabric drying technique offer new opportunities for energy saving. In this technique, high-frequency mechanical vibrations generated by the ultrasonic transducer are utilized to atomize water from a fabric in the liquid form, which demonstrates great potential for reducing energy use and drying time of the fabric drying process. Here, for the first time, fabric drying kinetics under different direct-contact ultrasonic drying conditions were investigated experimentally and analytically. The drying processes of four kinds of fabrics were experimentally tested under different ultrasonic transducer vibration frequency (115, 135, and 155 kHz) and input power (1.2, 2.5, and 4.4 W) conditions. According to the experimental data, five different kinds of models were applied to quantify the drying kinetics of fabrics during direct-contact ultrasonic drying. The models not only incorporated the transducer parameters but also the parameters related to the nature of fabric. Our evaluation results of model prediction performance demonstrated that the two empirical models, i.e., the Weibull model and the Gaussian model, were superior to the three semi-theoretical models for anticipating the drying kinetics of fabrics under direct-contact ultrasonic drying. Furthermore, the Weibull model is more suitable for practical energy-efficient direct-contact ultrasonic fabric drying applications compared with the Gaussian model.

An empirical exploration of performance metrics for event detection algorithms in Non-Intrusive Load Monitoring

The field of Non-Intrusive Load Monitoring (NILM) gained prominence due to its promise of inferring the energy consumption of individual appliances by analyzing only the aggregated consumption. Still, despite some research efforts towards producing meaningful comparisons between approaches, it is not yet possible to find a proven and formally accepted set of metrics to do this. Against this background, this paper focuses on understanding the challenges of defining a consistent set of performance metrics for this problem. More concretely, it reports on an empirical exploration of 23 performance metrics’ behavior when applied to event-detection algorithms, identifying relationships and clusters between them. The results indicate that when applied to this problem, the performance metrics will show some considerable differences in behavior compared to other, more traditional, machine-learning domains. The results also suggest that most of the differences occur due to the unbalanced nature of the event detection problem, in which the number of positive cases (True Positives and True Negatives) is much higher than the number of false situations (False Positives and False Negatives). Furthermore, the results suggest that additional research is needed to find proper domain-specific performance metrics that take full consideration of the properties of the aggregated load.

Appliance-specific feedback and social comparisons: Evidence from a field experiment on energy conservation

The provision of feedback about individual electricity consumption is a widely used approach to promote pro-environmental behavior. This form of feedback typically invokes social comparisons by informing households about their aggregate electricity consumption relative to others. While previous research has shown that aggregate consumption feedback translates into significant energy savings, the potential for further reductions may remain untapped because households lack knowledge about their appliance energy consumption patterns. In this paper, we present evidence from a field experiment, where we provide residents with feedback about their electricity consumption, specific to a high-energy use appliance (i.e. air-conditioner). We provide the relevant social norm information by varying the reference group of each resident. We find that our appliance-specific feedback is a powerful tool to curb electricity consumption. Residents significantly reduce their average air-conditioning usage by 17% in our treatment groups. Notwithstanding, our effects are not driven by comparative feedback with respect to different reference groups. We interpret this as encouraging evidence to promote the use of appliance-specific feedback to realize energy savings.

Appliance use behavior modelling and evaluation in residential buildings: A case study of television energy use

Aside from the consumption for domestic cooling and heating, the energy consumption of appliances in a household is a large amount. This energy use by appliances is highly dependent on the activities of the occupants in a building. Occupant behavior in residential buildings possesses the characteristic of randomness, variety, and complexity. More specifically, the profiles of energy consumption and appliance use are highly dependent on the timing of residents’ activities. Thus, it is necessary to model the domestic energy use profile with high temporal resolution. For example, in the context of energy demand response analysis, it is crucial to take occupant behavior into consideration. This article presents a thorough and detailed method for generating an appliance use pattern based on measured power data in real households, taking the television as a case study. The study develops a stochastic model based on appliance use data at one-minute temporal resolution. The proposed model firstly extracts feature parameters to depict the use behavior of the appliance. Then, one stochastic simulation model is established with the input of feature parameters. An evaluation method for this appliance behavior model is also proposed. As a result of one case study simulating appliance consumption at a variety of scales of households, it was concluded that the model can enable applicability to further household energy profile simulation.

Non-intrusive energy disaggregation by detecting similarities in consumption patterns


 
 
 Breaking down the aggregated energy consumption into a detailed consumption per appliance is a crucial tool for energy efficiency in residential buildings. Non-intrusive load monitoring allows implementing this strategy using just a smart energy meter without installing extra hardware. The obtained information is critical to provide an accurate characterization of energy consumption in order to avoid an overload of the electric system, and also to elaborate special tariffs to reduce the electricity cost for users. This article presents an approach for energy consumption disaggregation in households, based on detecting similar consumption patterns from previously recorded labelled datasets. The experimental evaluation of the proposed method is performed over four different problem instances that model real household scenarios using data from an energy consumption repository. Experimental results are compared with twobuilt-in algorithms provided by the nilmtk framework (combinatorial optimization and factorial hidden Markov model). The proposed algorithm was able to achieve accurate results regarding standard prediction metrics. The accuracy was not affected in a significant manner by the presence of ambiguity between the energy consumption of different appliances or by the difference of consumption between training and test appliances.
 
 

A Machine Learning Approach for Energy-Efficient IoT Systems

The energy challenge in IoT refers to the significant energy consumption of IoT devices, which can lead to sustainability issues, shorter battery life, and increased operating costs. IoT devices are known for their high energy consumption, and optimizing their energy usage can have a significant impact on sustainability and cost. Machine learning (ML) can learn from data and patterns to predict and control energy consumption in IoT systems, making them more energy efficient. The main contribution of this paper is the establishment of a novel deep learning framework for enhanced predictive modeling of energy consumption in IoT networks to help realize Energy-efficient IoT systems. our framework applies recurrent processing to capture long-term relations in the energy consumption of IoT appliances. Then, the self-attention mechanism is devised to help the model to focus on important predictive features. Simulation experiments against the competing ML baselines demonstrate the predictive capability of our framework.

Blockchain Enabled Distributed Demand Side Management in Community Energy System With Smart Homes

Existing work in energy demand side management focuses on the interaction between the utility grid and consumers. However, the previous technique is not focused on energy trading in local community of a renewable energy generation, distributed demand side management and not suitable for real-time environment. This paper presents a distributed demand side management system among multiple homes in community microgrid, with the integration of the internet of things smart meter and in the presence of renewable energy sources. The proposed energy consumption game is formulated for minimizing the cost of electricity in the individual home and the total cost of energy consumption in the whole community. The smart home users are playing game by optimizing their own daily energy consumption of appliances. The multiple participants include the self renewable generation of users, shared community microgrid and optional utility company. Each participant applies its best strategy to minimize energy consumption cost and users can maintain their own privacy of energy consumption. Moreover, the proposed scheme is distributed on blockchain, which provides a trusted communication medium between the participants. It enforces the autonomous monitoring of smart appliances and the billing of electricity consumption via smart contracts. Solidity smart contract is deployed to facilitate the execution of transactions without the involvement of third party in the smart community. Comparison of the results show that the proposed approach minimizes the total cost of energy consumption as well as each user's energy consumption cost.

Prediction model of household appliance energy consumption based on machine learning

This paper presents and discusses the prediction model of household appliance energy consumption. The paper discusses data filtering to remove non-predictive parameters, and used for model training. Five prediction models are trained with repeated cross validation and evaluated in the testing set: support vector machine (SVM), k nearest neighbor (KNN), random forest (RF), extreme random forest (ERF), and long short-term memory network (LSTM). Among them, the LSTM, which is a deep learning method, behaves the best performance. LSTM has the highest R2 (0.97) and the lowest RMSE (21.36) in the testing set. The analysis results show that the deep learning method has advantages in the prediction of household appliance energy consumption.

Big Data Energy Management, Analytics and Visualization for Residential Areas

With the rapid development of IoT based home appliances, it has become a possibility that home owners share with Utilities in the management of home appliances energy consumption. Thus, the proposed work empowers home owners to manage their home appliances energy consumption and allow them to compare their consumption with respect to their local community total consumption. This serves as a nudge in consumer’s behavior to schedule their home appliances operation according to their local community consumption profile and trend. Utilizing the same common communication infrastructure, it also allows the utilities on different consumption levels (community, state, country) to monitor and visualize the energy consumption in their respective grid segments on daily, monthly, and yearly basis. A high-speed distributed computing cluster based on commodity hardware with efficient big data mathematical algorithm is employed in this work. To achieve this, two big data processing paradigms are evaluated with a set of qualitative and quantitative metrics with subsequent recommendations. One million smart meter data is simulated to access individual homes. With the utilization of distributed storage and computing cluster for handling energy big data, the utilities can perform consumer load analysis and visualization on a scale of one million consumers. This helps the utilities in providing consumers a more accurate representation of how much energy they are consuming with greater granularity and with respect to their local community. Consumer and Utility centric queries are developed to create a web-based real time energy consumption management system presented in terms of dashboard charts, graphs, and reports that can be accessed by the consumer and utility providers remotely.

Appliance Specific Feedback and Social Comparisons: Evidence From a Field Experiment on Electricity Saving

The provision of feedback about individual electricity consumption is a widely used approach to promote pro-environmental behavior. This form of feedback typically invokes social comparisons by informing households about their aggregate electricity consumption relative to others. While previous research has shown that aggregate consumption feedback translates into significant energy savings, the potential for further reductions may remain untapped because households lack knowledge about their appliance energy consumption patterns. In this paper, we present evidence from a field experiment, where we provide residents with feedback about their electricity consumption, specific to a high-energy use appliance (i.e. air-conditioner). We provide the relevant social norm information by varying the reference group of each resident. We find that our appliance-specific feedback is a powerful tool to curb electricity consumption. Residents significantly reduce their average air-conditioning usage by 17\% in our treatment groups. Notwithstanding, our effects are not driven by comparative feedback with respect to different reference groups. We interpret this as encouraging evidence to promote the use of appliance-specific feedback to realize energy savings.

Typical occupancy profiles and behaviors in residential buildings in the United States

The energy performance of a residential building is highly dependent on occupant's presence or non-presence in a building and their interactions with energy-consuming appliances. Typical occupancy schedules for residential buildings must be defined for applications such as building energy modeling as well as for assessing energy savings associated with the use of occupancy sensing technologies and occupancy-dependent controls. Currently, commonly used simulation programs assume a typical occupancy schedule, however, there is significant opportunity for improvement to these schedules as this is generally based on engineering judgement. This research uses 12 years of the American Time Use Survey (ATUS) data to develop typical occupancy schedules for a range of household types and occupant age ranges. This is compared to currently utilized residential occupancy schedules. In many cases the developed schedules exhibit similar patterns, however, differences are also found to be as high as 41% for certain periods of time. For occupancy sensing applications, the spatial-temporal distribution of occupants in residential buildings is also evaluated. These locations vary based on temporal factors as well as demographic factors such as age and number of occupants. The results of this research work towards improved occupancy schedule development can benefit both industry professionals and researchers.