CENTRE FOR INFORMATICS AND SYSTEMS OF THE UNIVERSITY OF COIMBRA

Funder

Organizational Unit

Publications

Designing robust forecasting ensembles of Data-Driven Models with a Multi-Objective Formulation: An application to Home Energy Management Systems

Publication . Ruano, Antonio; Ruano, Maria

This work proposes a procedure for the multi-objective design of a robust forecasting ensemble of data-driven models. Starting with a data-selection algorithm, a multi-objective genetic algorithm is then executed, performing topology and feature selection, as well as parameter estimation. From the set of non-dominated or preferential models, a smaller sub-set is chosen to form the ensemble. Prediction intervals for the ensemble are obtained using the covariance method. This procedure is illustrated in the design of four different models, required for energy management systems. Excellent results were obtained by this methodology, superseding the existing alternatives. Further research will incorporate a robustness criterion in MOGA, and will incorporate the prediction intervals in predictive control techniques.

2023-07-25Journal article

Open Access

Unsupervised eeg preictal interval identification in patients with drug-resistant epilepsy

Publication . Leal, Adriana; Curty, Juliana; Lopes, Fábio; Pinto, Mauro F.; Oliveira, Ana; Sales, Francisco; Bianchi, Anna M.; Ruano, Maria; Dourado, António; Henriques, Jorge; Teixeira, César A.

Typical seizure prediction models aim at discriminating interictal brain activity from pre-seizure electrographic patterns. Given the lack of a preictal clinical definition, a fixed interval is widely used to develop these models. Recent studies reporting preictal interval selection among a range of fixed intervals show inter- and intra-patient preictal interval variability, possibly reflecting the heterogeneity of the seizure generation process. Obtaining accurate labels of the preictal interval can be used to train supervised prediction models and, hence, avoid setting a fixed preictal interval for all seizures within the same patient. Unsupervised learning methods hold great promise for exploring preictal alterations on a seizure-specific scale. Multivariate and univariate linear and nonlinear features were extracted from scalp electroencephalography (EEG) signals collected from 41 patients with drug-resistant epilepsy undergoing presurgical monitoring. Nonlinear dimensionality reduction was performed for each group of features and each of the 226 seizures. We applied different clustering methods in searching for preictal clusters located until 2 h before the seizure onset. We identified preictal patterns in 90% of patients and 51% of the visually inspected seizures. The preictal clusters manifested a seizure-specific profile with varying duration (22.9 +/- 21.0 min) and starting time before seizure onset (47.6 +/- 27.3 min). Searching for preictal patterns on the EEG trace using unsupervised methods showed that it is possible to identify seizure-specific preictal signatures for some patients and some seizures within the same patient.

2023Journal article

Open Access

From home energy management systems to communities energy managers: the use of an intelligent aggregator in a community in Algarve, Portugal

Publication . Gomes, Isaías; Graca Ruano, Maria; Ruano, Antonio

This paper describes the development of community energy management systems (CEMS). A CEMS allows optimal energy sharing within energy communities, as it is a central system that makes the global management of the entire community. The proposed CEMS is based on mixed-integer linear programming (MILP), operating under the receding horizon concept of Model Predictive Control (MPC). A systematic classification of electric appliances, the use of external information such as weather information and energy prices, as well as the use of intelligent forecasting techniques enables the proposed approach to achieve an excellent efficiency. It also allows for an easy installation of as well as a smooth scaling with an increasing number of houses. The system is tested in a real community in Algarve, Portugal. Different simulations are compared to experimental operation and include cases with and without sharing of energy, different resources allocated to the houses considered, and the use of different tariffs. CEMS formulations include sharing of energy without restriction, as well as employing different allocation coefficients strategies. The results show that for the community under study when managed by CEMS such as the one presented in this paper, it would result in significant cost reductions when compared to the case where there is no energy community.

2023Journal article

Restricted