Open Access
Open access
Energy and AI, volume 15, pages 100329

Reconstructing hourly residential electrical load profiles for Renewable Energy Communities using non-intrusive machine learning techniques

Lorenzo Giannuzzo 1, 2
Francesco Demetrio Minuto 1, 2
Daniele Salvatore Schiera 1, 2
Andrea Lanzini 1, 2
1
 
Department of Energy (DENERG), Polytechnic of Turin, Corso Duca degli Abruzzi 24, 10129 Turin, Italy
2
 
Energy Center Lab, Polytechinc of Turin, via Paolo Borsellino 38/16, 10152 Turin, Italy
Publication typeJournal Article
Publication date2024-01-01
Journal: Energy and AI
scimago Q1
wos Q1
SJR2.160
CiteScore16.5
Impact factor9.6
ISSN26665468
General Energy
Artificial Intelligence
Engineering (miscellaneous)
Abstract
The successful implementation of Renewable Energy Communities (RECs) involves maximizing the self-consumption within a community, particularly in regulatory contexts in which shared energy is incentivized. In many countries, the absence of a metering infrastructure that provides data at an hourly or sub-hourly resolution level for low-voltage users (e.g., residential and commercial users) makes the design of a new energy community a challenging task. This study proposes a non-intrusive machine learning methodology that can be used to generate residential electrical consumption profiles at an hourly resolution level using only monthly consumption data (i.e., billed energy), with the aim of estimating the energy shared by RECs. The proposed methodology involves three phases: first, identifying the typical load patterns of residential users through k-Means clustering, then implementing a Random Forest algorithm, based on monthly energy bills, to identify typical load patterns and, finally, reconstructing the hourly electrical load profile through a data-driven rescaling procedure. The effectiveness of the proposed methodology has been evaluated through an REC case study composed by 37 residential users powered by a 70 kWp photovoltaic plant. The Normalized Mean Absolute Error (NMAE) and the Normalized Root Mean Squared Error (NRMSE) were evaluated over an entire year and whenever the energy was shared within the REC. The Relative Absolute Error was also measured when estimating the shared energy at both a monthly (MRAE) and at an annual basis. (RAE). A comparison between the REC load profile reconstructed using the proposed methodology and the real load profile yielded an overall NMAE of 20.04%, an NRMSE of 26.17%, and errors of 18.34% and 23.87% during shared energy timeframes, respectively. Furthermore, our model delivered relative absolute errors for the estimation of the shared energy at a monthly and annual scale of 8.31% and 0.12%, respectively.
Bianchi F.R., Bosio B., Conte F., Massucco S., Mosaico G., Natrella G., Saviozzi M.
Applied Energy scimago Q1 wos Q1
2023-03-01 citations by CoLab: 33 Abstract  
The use of reversible solid oxide cells within a renewable energy community results a promising application which permits to balance the temporal mismatch between renewable energy production and users’ demand through hydrogen as energy vector. Differently from batteries and supercapacitors, this technology is characterized by a high stored energy density and a negligible daily self-discharge. Nevertheless, the system management is more complex requiring cell behaviour optimization and hydrogen storage control. Here this work proposes a control algorithm for reversible solid oxide cell operation coupled to renewable energy sources within a renewable energy community formed by an aggregation of fifteen residential customers. Based on forecasts of loads and renewable energy production, the proposed algorithm, a stochastic model predictive control, can optimize system operation aiming at economic benefit maximization. The transition between the fuel cell mode for power generation and the electrolysis mode for energy storage through hydrogen production was set considering the available renewable energy, the power demand of community members and the energy sell-back price in order to increase the auto-consumed amount as well as to favour the electricity exchange within the renewable energy community. Since the reversible solid oxide cell is the key-point in such a system, SIMFC-SIMEC (SIMulation of Fuel Cells and Electrolysis Cells), a physically based 2D model, allowed an effective prediction of cell behaviour deriving the efficiency of electricity and hydrogen production from local physicochemical feature and working parameter gradients on each stacked cell plane.
Haji Bashi M., De Tommasi L., Le Cam A., Relaño L.S., Lyons P., Mundó J., Pandelieva-Dimova I., Schapp H., Loth-Babut K., Egger C., Camps M., Cassidy B., Angelov G., Stancioff C.E.
2023-02-01 citations by CoLab: 61 Abstract  
In this study, the European-level regulations governing energy communities are reviewed. This is important as recognizing definitions, rights, and obligations in the regulatory world, for the energy community actors, such as citizen/renewable energy communities, active consumers, and renewables self-consumers, is vital to energy community development. This enables evaluating the comprehensiveness of regulations for functionality of the realized technical counterparts, namely microgrids, energy hubs, virtual power plants, and prosumers. By mapping the energy community actors to realized technical counterparts, the key finding of this study is that for certain cases, a right, an activity sector, or a technical feature necessitates matching to an energy community actor with a broader activity domain. This brings unnecessary rights and obligations to the intended realized technical counterpart. However, technical features prevent claiming those rights or meeting additional obligations. The study is then extended to national legislation of selected European countries, i.e., Austria, Bulgaria, Germany, Ireland, Latvia, Poland, and Spain. Regulatory misalignments or improvements in comparison to the European legislation are also reported. To address benefits and barriers associated with energy communities, a survey on collective energy actors (e.g., municipalities, private sector suppliers, city councils, energy service companies, etc) involved in energy transition is conducted. The key finding of the survey is that rather than economic benefits, achieving renewable targets and participating in energy-based social activities such as improving the energy efficiency are the primary motivations behind energy community initiatives.
Mignoni N., Scarabaggio P., Carli R., Dotoli M.
Control Engineering Practice scimago Q1 wos Q1
2023-01-01 citations by CoLab: 32 Abstract  
Recently, the decreasing cost of storage technologies and the emergence of economy-driven mechanisms for energy exchange are contributing to the spread of energy communities. In this context, this paper aims at defining innovative transactive control frameworks for energy communities equipped with independent service-oriented energy storage systems. The addressed control problem consists in optimally scheduling the energy activities of a group of prosumers, characterized by their own demand and renewable generation, and a group of energy storage service providers, able to store the prosumers’ energy surplus and, subsequently, release it upon a fee payment. We propose two novel resolution algorithms based on a game theoretical control formulation, a coordinated and an uncoordinated one, which can be alternatively used depending on the underlying communication architecture of the grid. The two proposed approaches are validated through numerical simulations on realistic scenarios. Results show that the use of a particular framework does not alter fairness, at least at the community level, i.e., no participant in the groups of prosumers or providers can strongly benefit from changing its strategy while compromising others’ welfare. Lastly, the approaches are compared with a centralized control method showing better computational results. • Two innovative transactive control frameworks for energy storage services are defined. • The considered system includes prosumers and energy storage service providers. • The resolution algorithms are based on a game theoretical formulation. • The transactive control frameworks are numerically validated on realistic scenarios. • Results show the effectiveness and scalability of the proposed control approaches.
Zhou Y.
Energy and AI scimago Q1 wos Q1 Open Access
2022-11-01 citations by CoLab: 54 Abstract  
• AIapplications in energy supply, storage, district demands,and energy management. • Supervised and unsupervised learningforclassification,regressionand clustering. • Reinforcement learningfor online optimal scheduling and energy management. • Battery state-of-charge, lifetime estimation, fault detection and diagnosis with AI. • ML formicrogridadaptive control, smart energy trading and decision-marking. • AI applications in low-carbontransitionfrom renewable torenewable-supply-storage. Energy paradigm transition towards the carbon neutrality requires combined and continuous efforts in cleaner power production, advanced energy storages, flexible district energy demands and energy management strategies. Applications of cutting-edge machine learning techniques can improve the system reliability with advanced fault detection and diagnosis (FDD, automation with agent-based reinforcement learning, flexibility with model predictive controls, and so on. In this study, a comprehensive review on artificial intelligence applications in carbon-neutral district community, has been conducted, from perspectives of energy supply, energy storage, district demands and energy management. Classifications and underlying mechanisms on ML techniques have been demonstrated, including supervised, unsupervised, reinforcement and deep learning. Afterwards, practical applications of ML have been reviewed, in respect to renewable energy supply, hybrid energy storages, district energy demand and advanced energy management. Results indicate that, supervised learning was mainly applied in classification and regression, and unsupervised learning was mainly applied in clustering. The reinforcement learning is mainly applied in on-line optimal scheduling for building energy management. With respect to clean energy supply, ML in solar and wind energy systems mainly include solar irradiance forecasting, wind resource forecasting, PV power prediction, maximum power point tracking (MPPT) for smart control, fault detection and diagnosis. ML in fuel cells mainly includes performance prediction, material selection, combination and so on. Furthermore, in respect to hybrid energy storages, ML in electrochemical battery includes dynamic thermal/electrical behavior, battery sizing and optimization, state-of-charge prediction, battery lifetime estimation, fault detection and diagnosis analysis. ML in sensible energy storages mainly include load prediction and storage capacity sizing, dynamic scheduling for cost saving, thermal stratification analysis and dynamic performance prediction. Advances in energy management with ML mainly include dispatch on stochastic and intermittent renewable power, microgrid adaptive control, smart energy trading with controls and decision-marking. Research tendency over the recent past several years indicates that, critical areas for low-carbon energy systems transit from the only renewable systems (59.4% in 2016) towards both renewable energy supply and energy storages (35.1% and 34.1%, respectively), such as battery, capacitors/supercapacitors, sensible/latent heat storages, compressed air storage and hydrogen storage. This study can provide a holistic overview and in-depth thinking on artificial intelligence in the carbon-neutral district transition.
Liu Z., Sun Y., Xing C., Liu J., He Y., Zhou Y., Zhang G.
Energy and AI scimago Q1 wos Q1 Open Access
2022-11-01 citations by CoLab: 113 Abstract  
• Functional contributions of AI techniques for large-scale renewable energy integrations were discussed. • Practical applications and effectiveness of various AI techniques were analyzed. • Limitations and challenges associated with large-scale renewable energy integrations using AI techniques were summarized. • Some promising research perspectives and recommendations were proposed. The vigorous expansion of renewable energy as a substitute for fossil energy is the predominant route of action to achieve worldwide carbon neutrality. However, clean energy supplies in multi-energy building districts are still at the preliminary stages for energy paradigm transitions. In particular, technologies and methodologies for large-scale renewable energy integrations are still not sufficiently sophisticated, in terms of intelligent control management. Artificial intelligent (AI) techniques powered renewable energy systems can learn from bio-inspired lessons and provide power systems with intelligence. However, there are few in-depth dissections and deliberations on the roles of AI techniques for large-scale integrations of renewable energy and decarbonisation in multi-energy systems. This study summarizes the commonly used AI-related approaches and discusses their functional advantages when being applied in various renewable energy sectors, as well as their functional contribution to optimizing the operational control modalities of renewable energy and improving the overall operational effectiveness. This study also presents practical applications of various AI techniques in large-scale renewable energy integration systems, and analyzes their effectiveness through theoretical explanations and diverse case studies. In addition, this study introduces limitations and challenges associated with the large-scale renewable energy integrations for carbon neutrality transition using relevant AI techniques, and proposes further promising research perspectives and recommendations. This comprehensive review ignites advanced AI techniques for large-scale renewable integrations and provides valuable informational instructions and guidelines to different stakeholders (e.g., engineers, designers and scientists) for carbon neutrality transition.
Conte F., D’Antoni F., Natrella G., Merone M.
Energy and AI scimago Q1 wos Q1 Open Access
2022-11-01 citations by CoLab: 33 Abstract  
In this study, we propose a hybrid AI optimal method to improve the efficiency of energy management in a smart grid such as Renewable Energy Community. This method adopts a Time Delay Neural Network to forecast the future values of the energy features in the community. Then, these forecasts are used by a stochastic Model Predictive Control to optimize the community operations with a proper control strategy of Battery Energy Storage System. The results of the predictions performed on a public dataset with a prediction horizon of 24 h return a Mean Absolute Error of 1.60 kW, 2.15 kW, and 0.30 kW for photovoltaic generation, total energy consumption, and common services, respectively. The model predictive control fed with such predictions generates maximum income compared to the competitors. The total income is increased by 18.72% compared to utilizing the same management system without exploiting predictions from a forecasting method. • We develop a new hybrid AI method to optimally manage the energy resources within a renewable energy community. • We present a new method for efficient management of users’ demand and photovoltaic generation uncertainties. • We optimize a Time Delay Neural Network to forecast future solar generation and power demand. • We present a stochastic model predictive control to maximize income through battery management. • The method is tested on a public dataset, achieving an increase of the community income by up to 18.72%
Volpato G., Carraro G., Cont M., Danieli P., Rech S., Lazzaretto A.
Energy scimago Q1 wos Q1
2022-11-01 citations by CoLab: 50 Abstract  
Energy communities are regulatory tools promoting aggregations of users to foster the shift towards a renewable distributed generation. First in the literature, this paper addresses together three main aspects affecting the convenience of these aggregations: the complementarity between generation and demand of different prosumers, the criterion allocating the operating costs of energy communities, and the application of demand-response programs. The goal is quantifying the relative weight of these aspects using Mixed-Integer Linear Programming to minimize the operating costs of citizen and renewable energy communities, where prosumers are connected to the grid as single entity, or separately. Incentive- or price-based demand-response programs and a novel cost allocation criterion, which rewards the members with the highest economic benefit in passing from simple consumers to prosumers, are applied to each community configuration. Results allow identifying general guidelines for the optimal economic operation of energy communities: i) complementarity may reduce costs by 15–20%, ii) a fairer cost allocation criterion may reduce the bills of prosumers using free-of-charge renewables by 20–30% compared to those using dispatchable sources, and iii) price-based demand-response may reduce community costs beyond 50%. Eventually, directions of further research, as the impact of energy communities on a national power system, are drawn. • Three main aspects affect the aggregation of prosumers in energy communities. • Optimal complementarity of prosumers allows to reduce costs in the range 15–20%. • A novel cost allocation criterion for energy communities is proposed. • Price-based demand response reduces the renewable energy community costs up to 60%. • Citizen energy community has up to 39% lower costs than the renewable energy one.
Fang M., Xiang Y., Xu B., Wang T., Pan L., Liu Y., Liu J.
2022-11-01 citations by CoLab: 12 Abstract  
Massive residential power consumption information provides data support for the mining and analysis of load patterns. This article proposes a complete framework for load pattern identification, which mainly includes the clustering module and the classification module. Considering that the high-dimensional load profiling dataset will bring a heavy computational burden, multiple dimensional scaling is introduced in the process of data preprocessing. Then, an innovative mixture model based on regular vine copula mixture model (RVMM) is adopted for clustering typical load patterns. Finally, a random forest (RF) classifier constructed with certain load characteristic indexes and RVMM clustering results is employed as a supervised classification model to predict the categories of subsequent new customers, and the accuracy is calculated by the 10-fold cross-validation. It is demonstrated in the case study that the proposed RVMM algorithm exhibits better performance in the clustering validity evaluation. Besides, higher accuracy is achieved by the RF classifier.
Minuto F.D., Lanzini A.
2022-10-01 citations by CoLab: 57 Abstract  
Renewable Energy Communities are aggregations of final users/prosumers whose energy comes from renewable power generators. This work systematically assesses the rules of the current and new approaches to sharing the generated benefits obtainable from renewable energy production and its consumption among members of a virtual net-metering energy community. In this work, three sharing mechanism algorithms are proposed to distribute the energy community net profit among the members, utilizing a Token currency, in proportion to the service they provide: financing service, self-consumption service, and both financing and self-consumption services. An energy community, comprising of 100 households and supplied by a 100 kWp photovoltaic system, has been simulated. Six scenarios were designed to determine how profits could be shared between users and prosumers. We focused on the impact that the adoption of a specific sharing mechanism could have on the different categories of members (e.g., low-energy users, high-energy users with ownership, etc.). • Simulation of a virtual net-metering energy community. • The redistribution of profits among community members based on different ownership and sharing strategies. • A fair remuneration of profits is influenced significantly by the adopted sharing mechanisms. • The temporal load profile of users has a huge impact on the redistribution of profits.
Wang Q., Ding Y., Kong X., Tian Z., Xu L., He Q.
Energy scimago Q1 wos Q1
2022-09-01 citations by CoLab: 7 Abstract  
Air conditioning systems are generally considered to have the greatest flexibility potential in buildings that can be flexibly regulated with thermal storage to reduce the interaction with the power grid and increase demand response benefits. In previous studies, the flexibility of air-conditioning systems was reflected through time-of-use tariffs. However, a strategy that only factors the tariffs incurs a greater operational energy consumption. In this study, a flexibility factor was established and incorporated into the multi-objective optimization process, together with the operational energy consumption, as two optimization objectives. After obtaining typical load patterns using a two-step clustering method, for multi-objective decision-making in the day-ahead operation, the entropy-grey technique for order preference by similarity to an ideal solution method is used. Considering an office building as a case study, we found that the optimized flexibility factor can reach 0.31 and 0.99 during a week of operation in winter and summer, on average, respectively, and achieved a cumulative energy-saving effect of 17.98% and 35.49%. In addition, the two-step clustering method can better demonstrate the flexibility factor than the single-step clustering. • Typical air conditioning load patterns are selected by a two-step clustering method. • Detailed load boundary division leads to better energy-saving operation strategies. • Flexibility factor is incorporated into the multi-objective optimization process. • Demand response benefits are fully realized by balancing cost and efficiency.
Köhler S., Rongstock R., Hein M., Eicker U.
Energy and Buildings scimago Q1 wos Q1
2022-09-01 citations by CoLab: 13 Abstract  
The simulation of residential electricity load profiles (ELPs) has always played an important role for designing and evaluating energy systems for buildings or entire neighborhoods. Large-scale measurement data, the counterpart to these synthetic data, are often not available or only available at great expense in terms of time and under consideration of data protection. Therefore, sometimes very detailed and elaborate load profile generators are created, which allow the simulation of different scenarios even without measured data. Simulating electricity load profileson a large scale on the one hand, and as detailed as possible on the other, is subject to several challenges. A particular challenge is the assessment of representativeness and the question of which measures are used to evaluate this. Specifically, which measures indicate whether the curve progression of a synthetic load curve becomes more similar to measured curves and when it does not. Electricity load profiles are highly complex structures that depend on numerous conditions. This paper aims to present an approach that addresses the issue of assessing the similarity or representativeness of electricity load profiles. Emphasis is placed on the comparative measures that are expected to indicate the representativeness or similarity between synthetic and measured electricity load profile data. In the first step, comparative measures used in the literature are gathered as well as classified with respect to their statement on the similarity of electricity load profiles. It is of essence, that similarity in this paper corresponds to the likeness and not the sameness of electricity load profile data. Adding to the measures found in the literature, three further similarity measures are introduced. Using measured electricity load profile data from a case study and synthetic electricity load profile data from three different load profile generators, selected similarity measures are calculated and compared. It is found that in addition to measures of position, central tendency and dispersion, the newly introduced complexity measures may substantiate the expressiveness with respect to the similarity of electricity load profiles. In particular, the complexity measure of the fractal dimension seems to be a potential for further similarity studies.
Weckesser T., Dominković D.F., Blomgren E.M., Schledorn A., Madsen H.
Applied Energy scimago Q1 wos Q1
2021-11-01 citations by CoLab: 85 Abstract  
In this paper, an extensive study of Renewable Energy Communities and their potential impact on the electric distribution grid has been carried out. For that purpose, a Linear Programming optimization model sizing the energy community’s Photo-Voltaic and Battery Energy Storage System was developed. The linear programming model was soft coupled with power flow analysis to investigate the impact of different energy community configurations on distribution grids. Different distribution grids (city, suburban, village), different energy community configuration, different operating strategies and different battery placements were investigated. The results showed that when the battery is located at the beginning of the feeder, then the energy community does not impact the observed minimum and maximum voltage. Moreover, it was found that depending on the energy community’s operating strategy the low-voltage grid loading can be reduced by up to 58 %. The energy community’s sizing showed that optimal capacities of photo-voltaics and communal batteries were up to three times larger for the case of city grid, following the operating strategy of maximizing the energy community’s own economic benefit than in other operating strategies and grid types. • An extensive study of Renewable Energy Communities (ECs) & their impact on the grid. • An optimization model for sizing PV&BESS considering different operation strategies. • Grid impact analysis for different distribution grid types and EC configurations. • Identification of crucial EC properties, which determine its impact on the grid.
Zapata Castillo V., De Boer H., Maícas Muñoz R., Gernaat D.E., Benders R., van Vuuren D.
2021-10-10 citations by CoLab: 2
Hanke F., Guyet R., Feenstra M.
2021-10-01 citations by CoLab: 129 Abstract  
A growing energy justice literature underlines that complex energy injustices in energy transition disproportionally affect vulnerable and energy-poor households. Literature and policies discuss renewable energy communities’ (RECs) potential to enable citizen participation in energy transition and shape a just transition. Low-income and energy-poor households could benefit from granting access to affordable energy tariffs and energy efficiency measures when participating in RECs. Recent EU legislation highlights RECs’ social role in energy poverty alleviation and stipulates the participation of all social groups in RECs, especially those groups that are underrepresented under RECs’ members. In this light, the energy justice framework is increasingly applied to analyse RECs’ social contributions in different countries. Still, empirical evidence of RECs’ capacity to include underrepresented and vulnerable groups and mitigate energy poverty as a particular form of energy justice remains scarce. Drawing on data collected among 71 European RECs, our exploratory research investigates how RECs engage in this social role by improving participatory procedures to enable vulnerable groups’ participation and by distributing affordable energy and energy efficiency to vulnerable households. Using the energy justice framework, we explore how RECs resonate with the three energy justice tenets (distributive, recognitional and procedural) by addressing underrepresented groups and energy poverty.
Saraf A., Kowli A.
2024-10-19 citations by CoLab: 0 Abstract  
Consumer load profiling involves examining patterns of energy consumption using available data. With smart meter data available at (sub-)hourly intervals, it is possible to use the it to generate daily load profiles that capture the typical consumption behavior across a representative day. Previous work has shown how fine-grained smart meter data and monthly consumption data separated by time-of-use can be translated into representative load profiles for a given group of consumers. In this work, an approach for load profile generation is proposed which only uses monthly energy consumption data without breaking it down further based on time-of-use such as peak, off-peak and mid-peak hours. Performance of data-driven models using random forest, XGboost and multi-layer perceptron is studied by reconstructing the load profiles on a known smart meter dataset. A comprehensive assessment of how the shape of the reconstructed load profile compares with the actual profile is provided by evaluating amplitude errors, time shifts and slope variations. Our investigations provide insights on the role of the cluster size on load profile reconstruction: predictions of profiles for large cluster sizes show improved accuracy in terms root mean-square errors and less bias. By gaining insights into when and how energy is consumed, utilities can implement measures to reduce costs, mitigate peak demand and enhance overall energy efficiency.
Stolte M., Minuto F.D., Lanzini A.
2024-08-01 citations by CoLab: 5 Abstract  
This article analyzes a power-to-hydrogen system, designed to provide high-temperature heat to hard-to-abate industries. We leverage on a geospatial analysis for wind and solar availability and different industrial demand profiles with the aim to identify the ideal sizing of plant components and the resulting Levelized Cost of Hydrogen (LCOH). We assess the carbon intensity of the produced hydrogen, especially when grid electricity is utilized. A methodology is developed to size and optimize the PV and wind energy capacity, the electrolyzer unit, and hybrid storage, by combining compressed hydrogen storage with lithium-ion batteries. The hydrogen demand profile is generated synthetically, thus allowing different industrial consumption profiles to be investigated. The LCOH in a baseline scenario ranges from 3.5 to 8.9 €/kg, with the lowest values in wind-rich climates. Solar PV only plays a role in locations with high PV full-load hours. It was found that optimal hydrogen storage can cover the users' demand for 2–3 days. Most of the considered scenarios comply with the emission intensity thresholds set by the EU. A sensitivity analysis reveals that a lower variability of the demand profile is associated with cost savings. An ideally constant demand profile results in a cost reduction of approximately 11 %.
Sedaghat A., Kalbasi R., Narayanan R., Mehdizadeh A., Soleimani S.M., Ashtian Malayer M., Iyad Al-Khiami M., Salem H., Hussam W.K., Sabati M., Rasul M., Masud K. Khan M.
Solar Energy scimago Q1 wos Q2
2024-07-01 citations by CoLab: 6
Minuto F.D., Crosato M., Schiera D.S., Borchiellini R., Lanzini A.
Energy Reports scimago Q2 wos Q2 Open Access
2024-06-01 citations by CoLab: 2 Abstract  
Renewable energy communities (RECs) offer a promising perspective for decarbonizing the building sector. This is accomplished by enhancing the uptake, among others, of citizen-owned rooftop photovoltaic systems. A key challenge lies in ensuring that photovoltaic generation matches the needs of community members, i.e. maximizing shared energy index. Indeed, the shared energy depends on the consumption habits of individual members and the rooftop characteristics, such as orientation and inclination of available pitches, which influence the production curve. Therefore, clear guidelines on which roof pitches are most suitable for PV generation within RECs might be helpful during the design of such communities. In this paper, we investigate the optimal orientations and tilt angles for PV systems in REC design. We conducted a robust Monte Carlo simulation of an energy community comprising 60 users, 30 of which are equipped with rooftop PV systems for a total of 150 kWp installed. Our analysis revealed that pitches with West and East offer comparable, if not better, shared energy values than those South-facing, consequently mitigating peak power dispatched to the grid. Besides, shared energy remains quite constant across various tilt angles. These findings suggest that buildings with non-South-facing roofs should not be overlooked, but embraced in the design of renewable energy communities as they can contribute significantly to shared energy.
Nutakki M., Mandava S.
2024-05-11 citations by CoLab: 0 PDF Abstract  
AbstractThe integration of smart homes into smart grids presents numerous challenges, particularly in managing energy consumption efficiently. Non-intrusive load management (NILM) has emerged as a viable solution for optimizing energy usage. However, as smart grids incorporate more distributed energy resources, the complexity of demand-side management and energy optimization escalates. Various techniques have been proposed to address these challenges, but the evolving grid necessitates intelligent optimization strategies. This article explores the potential of data-driven NILM (DNILM) by leveraging multiple machine learning algorithms and neural network architectures for appliance state monitoring and predicting future energy consumption. It underscores the significance of intelligent optimization techniques in enhancing prediction accuracy. The article compares several data-driven mechanisms, including decision trees, sequence-to-point models, denoising autoencoders, recurrent neural networks, long short-term memory, and gated recurrent unit models. Furthermore, the article categorizes different forms of NILM and discusses the impact of calibration and load division. A detailed comparative analysis is conducted using evaluation metrics such as root-mean-square error, mean absolute error, and accuracy for each method. The proposed DNILM approach is implemented using Python 3.10.5 on the REDD dataset, demonstrating its effectiveness in addressing the complexities of energy optimization in smart grid environments.
Shabbir N., Vassiljeva K., Nourollahi Hokmabad H., Husev O., Petlenkov E., Belikov J.
Electronics (Switzerland) scimago Q2 wos Q2 Open Access
2024-04-09 citations by CoLab: 8 PDF Abstract  
Non-intrusive load monitoring (NILM) has emerged as a pivotal technology in energy management applications by enabling precise monitoring of individual appliance energy consumption without the requirements of intrusive sensors or smart meters. In this technique, the load disaggregation for the individual device is accrued by the recognition of their current signals by employing machine learning (ML) methods. This research paper conducts a comprehensive comparative analysis of various ML techniques applied to NILM, aiming to identify the most effective methodologies for accurate load disaggregation. The study employs a diverse dataset comprising high-resolution electricity consumption data collected from an Estonian household. The ML algorithms, including deep neural networks based on long short-term memory networks (LSTM), extreme gradient boost (XgBoost), logistic regression (LR), and dynamic time warping with K-nearest neighbor (DTW-KNN) are implemented and evaluated for their performance in load disaggregation. Key evaluation metrics such as accuracy, precision, recall, and F1 score are utilized to assess the effectiveness of each technique in capturing the nuanced energy consumption patterns of diverse appliances. Results indicate that the XgBoost-based model demonstrates superior performance in accurately identifying and disaggregating individual loads from aggregated energy consumption data. Insights derived from this research contribute to the optimization of NILM techniques for real-world applications, facilitating enhanced energy efficiency and informed decision-making in smart grid environments.

Top-30

Journals

1
1

Publishers

1
2
3
4
5
6
7
1
2
3
4
5
6
7
  • We do not take into account publications without a DOI.
  • Statistics recalculated only for publications connected to researchers, organizations and labs registered on the platform.
  • Statistics recalculated weekly.

Are you a researcher?

Create a profile to get free access to personal recommendations for colleagues and new articles.
Share
Cite this
GOST | RIS | BibTex
Found error?