Bayesian hidden Markov models for performance-based regulation of continuity of electricity supply

Keywords

Statistical learning
Code:
20/2009
Title:
Bayesian hidden Markov models for performance-based regulation of continuity of electricity supply
Date:
Saturday 25th July 2009
Author(s):
Accoto, Nadia; Rydén, Tobias; secchi, Piercesare
Download link:
Abstract:
A fundamental aspect in the regulation of continuity of electricity supply is the identification of faults that could be caused by an exceptional event and therefore that are outside the utility control and responsibility. Different methods have been proposed during the years; the interpretation of the observed faults as a signal of an underlying system naturally leads to the analysis of the problem by means of an hidden Markov model. These models, in fact, are widely used for introducing dependence in data and/or for modelling observed phenomena depending on hidden processes. The application of this method shows that the model is able to identify exceptional events; moreover the study of the estimated model parameters gives rise to reality-linked considerations.