A New Framework for Multivariate Time Series Forecasting in Energy Management System

Uremovic, Niko; Bizjak, Marko; Sukic, Primoz; Stumberger, Gorazd; Žalik, Borut; Lukač, Niko

doi:10.1109/tsg.2022.3224559

Cited by 2 publications

(4 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Gaussian filter was used to preprocess the data in the offline setting, and Kalman filter in the online (real-time forecasting) setting. We provided the in-depth rationale and analysis of their respective uses and conducted an experiment determining the optimal parametrization in [6]. The effect of the filtering can be seen in Fig 1 . in the graph on the bottom left.…”

Section: Data Preprocessing Through Filteringmentioning

confidence: 99%

“…LSTM is a type of a recurrent neural network (RNN), which are specifically designed to model the temporal correlation within the timeseries data. We described the proposed architecture, training details and conducted experiments determining the optimal hyperparametrization of the forecast model and the training process in [6].…”

Section: Prediction Model Architecturementioning

confidence: 99%

“…Furthermore, we have developed an algorithm of selecting this smaller subset of samples from historical data, which further improved the forecasting accuracy both on old and new patterns, compared to the random selection. An in-depth description of the algorithm and the evaluation of adaptation approaches is presented in [6].…”

Section: Prediction Model Self-adaptationmentioning

confidence: 99%

“…Thus, the prediction models must be capable of selfestablishing and self-adaptation. The following section provides a deeper insight into a process that can give prediction models self-establishing and self-adaptation capability [6].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Prediction models and their role in advanced energy management systems supporting energy flexibility services

Uremovic¹,

Lukač²,

Sukič³

et al. 2023

RE&PQJ

View full text Add to dashboard Cite

Automated decision tools, such as advanced energy management systems, are required to involve the electrical grid users in energy flexibility services. This paper focuses on the prediction models as a substantial part of decision strategy in advanced energy management systems and on advanced energy management systems as a tool that supports the active involvement of electrical grid users in energy flexibility services. Prediction models' desired properties are self-establishing and self-adaptation, which require new solutions in data selection, filtering, processing and model learning. Some of these properties are investigated within this paper.

show abstract

Section: Data Preprocessing Through Filteringmentioning

confidence: 99%

Section: Prediction Model Architecturementioning

confidence: 99%

Section: Prediction Model Self-adaptationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations