Cookies Policy
The website need some cookies and similar means to function. If you permit us, we will use those means to collect data on your visits for aggregated statistics to improve our service. Find out More
Accept Reject
  • Menu
Publications

Publications by Manuel Matos

2023

Distributed Network-Constrained P2P Community-Based Market for Distribution Networks

Authors
Oliveira, C; Simoes, M; Bitencourt, L; Soares, T; Matos, MA;

Publication
ENERGIES

Abstract
Energy communities have been designed to empower consumers while maximizing the self-consumption of local renewable energy sources (RESs). Their presence in distribution systems can result in strong modifications in the operation and management of such systems, moving from a centralized operation to a distributed one. In this scope, this work proposes a distributed community-based local energy market that aims at minimizing the costs of each community member, accounting for the technical network constraints. The alternating direction method of multipliers (ADMM) is adopted to distribute the market, and preserve, as much as possible, the privacy of the prosumers' assets, production, and demand. The proposed method is tested on a 10-bus medium voltage radial distribution network, in which each node contains a large prosumer, and the relaxed branch flow model is adopted to model the optimization problem. The market framework is proposed and modeled in a centralized and distributed fashion. Market clearing on a day-ahead basis is carried out taking into account actual energy exchanges, as generation from renewable sources is uncertain. The comparison between the centralized and distributed ADMM approach shows an 0.098% error for the nodes' voltages. The integrated OPF in the community-based market is a computational burden that increases the resolution of the market dispatch problem by about eight times the computation time, from 200.7 s (without OPF) to 1670.2 s. An important conclusion is that the proposed market structure guarantees that P2P exchanges avoid the violation of the network constraints, and ensures that community agents' can still benefit from the community-based architecture advantages.

2023

Including Dynamic Security Constraints in Isolated Power Systems Unit Commitment/Economic Dispatch: a Machine Learning-based Approach

Authors
de Sousa, RP; Moreira, C; Carvalho, L; Matos, M;

Publication
2023 IEEE BELGRADE POWERTECH

Abstract
Isolated power systems with high shares of renewables can require additional inertia as a complementary resource to assure the system operation in a dynamic safe region. This paper presents a methodology for the day-ahead Unit Commitment/ Economic Dispatch (UC/ED) for low-inertia power systems including dynamic security constraints for key frequency indicators computed by an Artificial Neural-Network (ANN)-supported Dynamic Security Assessment (DSA) tool. The ANN-supported DSA tool infers the system dynamic performance with respect to key frequency indicators following critical disturbances and computes the additional synchronous inertia that brings the system back to its dynamic security region, by dispatching Synchronous Condensers (SC) if required. The results demonstrate the effectiveness of the methodology proposed by enabling the system operation within safe frequency margins for a set of high relevance fault type contingencies while minimizing the additional costs associated with the SC operation.

2023

Full distributed P2P market and distribution network operation based on ADMM: Testing and evaluation

Authors
Oliveira, C; Simoes, M; Soares, T; Matos, MA; Bitencourt, L;

Publication
2023 19TH INTERNATIONAL CONFERENCE ON THE EUROPEAN ENERGY MARKET, EEM

Abstract
This work models a distributed community-based market with diverse assets (photovoltaic generators and energy storage systems), accounting for network constraints and adopting the relaxed branch flow model. The market is modeled in a single and fully distributed approach, employing the alternating direction method of multipliers (ADMM) to prevent voltage and line capacity problems in the community network and improve data privacy and reduce the communication burden. Different scenarios, based on the penalty term and the agents' number, are tested to study the efficiency of the algorithm and the convergence rate of the ADMM distributed model. The proposed method is tested on 10-bus, 22-bus, and 33-bus medium voltage radial distribution networks, where each node contains a large prosumer with one or several assets. One important conclusion is that the implemented residual balancing technique improves the efficiency of the ADMM distributed algorithm by increasing the convergence rate and reducing the computational time.

1995

Electric distribution systems planning with fuzzy loads

Authors
Matos M.A.; Ponce de Leão M.T.;

Publication
International Transactions in Operational Research

Abstract
Distribution systems planning heavily depends on the predicted future consumptions in the service area. When statistical data exist about past consumptions, probabilistic forecasting methods may be applied, and expected cost/benifit and risk analysis are used to decide between different solutions. In most cases, however, this strategy is not applicable, due mainly to the lack of significant data (new developing areas, rapidly changing situations) and uncertainty of economic and social factors. In the latter case, the use of fuzzy models is an interesting alternative, accommodating expert planner's qualitative judgments about future loads and allowing us to use 'typical' load diagrams in new areas. The paper discusses the main concepts of electric distribution system planning when loads are fuzzy modeled, and presents an illustrative application example. © 1995.

2007

Forecasting Portugal global load with artificial neural networks

Authors
Fidalgo, JN; Matos, MA;

Publication
Artificial Neural Networks - ICANN 2007, Pt 2, Proceedings

Abstract
This paper describes a research where the main goal was to predict the future values of a time series of the hourly demand of Portugal global electricity consumption in the following day. In a preliminary phase several regression techniques were experimented: K Nearest Neighbors, Multiple Linear Regression, Projection Pursuit Regression, Regression Trees, Multivariate Adaptive Regression Splines and Artificial Neural Networks (ANN). Having the best results been achieved with ANN, this technique was selected as the primary tool for the load forecasting process. The prediction for holidays and days following holidays is analyzed and dealt with. Temperature significance on consumption level is also studied. Results attained support the adopted approach.

1997

Assessing error bars in distribution load curve estimation

Authors
Fidalgo, JN; Matos, MA; Ponce De Leao, MT;

Publication
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Abstract
Electrical distribution utilities have been dealing with the problem of estimation of distribution network load diagrams, either for operation studies or in forecasting models for planning purposes. Load curve assessment is essential for an efficient management of electric distribution systems. However, the only information available for most of the loads (namely LV loads) is related to monthly energy consumption. The general procedure uses measurements in consumers to construct inference engines that predict load curves using commercial information. This paper presents a new approach for this problem, based on Kohonen maps and Artificial Neural Networks (ANN) to estimate load diagrams for the Portuguese distribution utilities. A method for estimating error bars is also proposed in order to provide a high order information about the performance of load curve estimation process. Performance attained is discussed as well as the method to achieve confidence intervals of the main predicted diagrams. © Springer-Verlag Berlin Heidelberg 1997.

  • 10
  • 26