Home
IEEE MELECON 2020
MELECON 2020 - 4.4 - P1

Technical Session

IEEE Mediterranean Eletrotechnical Conference 2020

IEEE MELECON 2020

CONFERENCE HOME SCHEDULE TRACKS

Conference Track

Session 4.4 - PART I - Demand response and prosumers' aggregation

Not Authorized

You don't have privileges to see this content.

Learn more about how to register for an individual account.

GET ACCESS

Please Login

You should be logged in to see this content.

Don't you have an account?
Learn more about how to register for an individual account.

GET ACCESS

PRESENTATIONS
SESSION DETAILS

INVITED - Contribution of smart prosumers to the grid reliability

Diego Arnone - Engineering I.I. S.p.A., Italy

PREVIEW e-BADGE PRESENTATION

INVITED - Contribution of smart prosumers to the grid reliability

ABSTRACT

The power grid is continuously evolving toward the Smart Grid concept: data and electricity move together, and Information Technology plays an even more central role in this emerging panorama. Big and small energy producers as well as consumers become both reactive and proactive stakeholders that must communicate each other and cooperate by finding a compromise between their own strategic objectives and the maximization of the power grid reliability. Demand Response, prosumers aggregation, smart storage, energy sector coupling technologies are only some for the solutions under development to mitigate the power grid voltage and frequency fluctuations. Engineering I.I. S.p.A. is a big Italian IT company, founded 40 years ago, that gets more than one forth of its revenues from the Energy and Utilities domain. Thanks to huge investments in research and development activities, Engineering is actively participating to more than 100 research co-funded initiatives, both national and H2020 research projects. A significant number of these projects are focused on the design and development of innovative IT solutions for the power grid, the energy producers and the consumers. In the role of key technology provider, Engineering has taken part of more than 20 research projects concerning the collaboration between power grid and smart prosumers and this allows the company to offer reliable solutions to big customers like TERNA, ENEL, IREN and many other players in the energy domain. The speech will be an overview of a small subset of the “smart energy projects” that are currently ongoing in the Engineering Research and Development Laboratory.

Dynamic Coalitions of Prosumers in Virtual Power Plants for Energy Trading and Profit Optimization

Giuseppe Raveduto - Engineering Ingegneria Informatica S.p.A., Italy

PREVIEW e-BADGE PRESENTATION

Dynamic Coalitions of Prosumers in Virtual Power Plants for Energy Trading and Profit Optimization

ABSTRACT

Energy systems are becoming more oriented to a local scale, combining the self-provision of local communities with the delivery of services to the main grid. Distributed approaches are being developed to facilitate the transition towards decentralized, community-driven energy systems which contribute to the integration of renewables in existing power systems. In this paper, we present results from the ongoing research project eDREAM, which aims to provide flexibility on grid operation, support the integration of renewables, and develop strategies to manage energy storage systems. In particular, we propose an optimization problem tailored to optimize profits from energy trading in Virtual Power Plants (VPPs). A prototype has been implemented and it proved to be effective in optimizing the prosumers' pool and selecting the optimal subset for the coalition, taking as input the energy price during the time interval in which the coalition is being constructed.

Implementation of a Management System for Prosumer Energy Storage Scheduling in Smart Grids

Giuseppe Caravello - University of Palermo, Italy

PREVIEW e-BADGE PRESENTATION

Implementation of a Management System for Prosumer Energy Storage Scheduling in Smart Grids

ABSTRACT

Scheduling of energy storage systems of distributed generation is a key element for optimal renewable energy sources (RES) exploitation in smart grids, demand side response strategies implementation and prosumers interaction with Distribution System Operators (DSOs). In this context, this paper presents the implementation of a management system for energy storage scheduling, which allows taking appropriate account of the storage system behavior in real cases. To properly consider the real system operation, the employed battery model is modified, in order to include also the hysteresis phenomena in state of charge estimation in both charging and discharging operation. The validation of the proposed methodology is supported by experimental tests which allow comparing two different battery models. The obtained results show that better results can be obtained with the model including the hysteresis effect.

Ensemble learning with time-series clustering for aggregated short-term load forecasting

Petar Sarajcev - University of Split, Croatia

PREVIEW e-BADGE PRESENTATION

Ensemble learning with time-series clustering for aggregated short-term load forecasting

ABSTRACT

Load forecasting, as one of the important research areas of the smart grids, spans a wide range of methods, from traditional time-series econometric analyses to different machine learning and, recently, even deep learning approaches. This paper proposes a novel machine learning approach for short-term load time-series forecasting, which utilizes aggregate load clustering with ensemble learning based on the windowing method. Ensemble of base learners, comprised of gradient boosting, support vector machine (SVM) and random forest, is created by stacking models with an “elastic net” linear regression. Models hyper-parameters are fine-tuned using a grid search with cross-validation approach, except for the SVM, where Bayesian optimization is introduced. Features engineering and selection based on the importance analysis is employed, using weather and load time-series data. The mean absolute percentage error is used for verification. Obtained results show that the proposed approach exhibits accurate and robust predictions.

Unsupervised learning procedure for NILM applications

Gilles Jacobs - Université Libre de Bruxelles, Belgium

PREVIEW e-BADGE PRESENTATION

Unsupervised learning procedure for NILM applications

ABSTRACT

In a domestic context, NILM applications (Non-Intrusive Load Monitoring) allow users to know their electric consumption per appliance without having to install sensors for each appliance in their house. The aim of this paper is to present a new generic method allowing to discover the different appliances present in a house. Results obtained in this paper relies on supervised data obtained by sub-metering at the level of each appliance for 3 real domestic houses.

DEMAND Project: An algorithm for the assessment of the prosumers’ flexibility

Michele Cacioppo - University of Palermo, Italy

PREVIEW e-BADGE PRESENTATION

DEMAND Project: An algorithm for the assessment of the prosumers’ flexibility

ABSTRACT

Demand side aggregation represents an important opportunity for ancillary services provision due to the potential that the coordinated management of distributed resources has on affecting power systems’ operation. In this framework, the Aggregator has a main role and takes on different relationships with the other actors of the power system, usually becoming a mediator between the prosumers and the distribution system operator. The DEMAND project introduces a new point of view in demand side aggregation by proposing a new framework where the Aggregator is no more needed and prosumers can share and combine their flexibility in a new aggregation platform called Virtual Aggregation Environment (VAE). In this paper, an algorithm for the dayahead estimation of each prosumer’s flexibility is presented. The algorithm evaluates various couples of power variation and desired remuneration to send to the VAE for further elaborations. A case study shows how the algorithm works.

Investigation of autoregressive forecasting models for market electricity price

Vladimir Shikhin - National Research University "MPEI", Russia

PREVIEW e-BADGE PRESENTATION

Investigation of autoregressive forecasting models for market electricity price

ABSTRACT

Forecast models of ARIMA-type been investigated in application to market electricity prices behavior prediction in the form of time series. Paper presents the results of the forecast achieved accuracy study for constructing forecast via autoregressive statistical models and their close derivatives. Rather detailed computational procedures presented and supplied with numerical results. Adequacy verification of forecast mathematical models with reference to historical natural data in the form of time series carried out with reference to numerical estimation of the standard error. The achieved accuracy level of the designed predictive models for electrical energy market were found through Belgorod region in European part of Russian coincides with published results over international energy markets in Europe, America and Australia. Comparative analysis and interpretation of mathematical models for prediction, both published and obtained in this work leads to the conclusion that increasing complexity of statistical autoregressive forecast models (complexity of structures, the number of unknown parameters, the combination of heterogeneous components, the introduction of correction coefficients) only in individual cases and slightly increases the prediction accuracy. It is concluded that essential step effect of the forecast accuracy can be obtain through composed modeling of dependent variable with reference to the most influenced factors, and the problem to be solved is the design of the aggregate model structure.

SESSION CHAIRS

Diego Arnone

Engineering I.I. S.p.A., Italy

e-BADGE

Francesco Grasso

University of Florence, Italy

e-BADGE

Technical Session

IEEE Mediterranean Eletrotechnical Conference 2020

IEEE MELECON 2020

Session 4.4 - PART I - Demand response and prosumers' aggregation

INVITED - Contribution of smart prosumers to the grid reliability

Dynamic Coalitions of Prosumers in Virtual Power Plants for Energy Trading and Profit Optimization

Implementation of a Management System for Prosumer Energy Storage Scheduling in Smart Grids

Ensemble learning with time-series clustering for aggregated short-term load forecasting

Unsupervised learning procedure for NILM applications

DEMAND Project: An algorithm for the assessment of the prosumers’ flexibility

Investigation of autoregressive forecasting models for market electricity price

SESSION CHAIRS

Diego Arnone

Francesco Grasso