Echo State Network (ESN) attracted significant interest in the research activities in last years. In this paper some application to industrial cases are presented, considering in particular the energy and manufacturing sectors. In particular, load forecasting is crucial for penetrations of renewable energy sources and extension of programs in the paradigm of smart grids. Feed-Forward Neural Network (FFNN) based techniques have been widely used in recent years and applied to predict the electric load with high accuracy. This research work is focused on the use and comparison of neural network approaches, i.e. FFNN and ESN, on a dataset related to industrial application. The results of both models are compared based on their accuracy through experimental measurements and suitably defined metrics.

Echo State Network Performance in Electrical and Industrial Applications

Grimaccia F.;Mussetta M.
2020-01-01

Abstract

Echo State Network (ESN) attracted significant interest in the research activities in last years. In this paper some application to industrial cases are presented, considering in particular the energy and manufacturing sectors. In particular, load forecasting is crucial for penetrations of renewable energy sources and extension of programs in the paradigm of smart grids. Feed-Forward Neural Network (FFNN) based techniques have been widely used in recent years and applied to predict the electric load with high accuracy. This research work is focused on the use and comparison of neural network approaches, i.e. FFNN and ESN, on a dataset related to industrial application. The results of both models are compared based on their accuracy through experimental measurements and suitably defined metrics.
2020
Proceedings of the International Joint Conference on Neural Networks
978-1-7281-6926-2
Demand Response programs
Echo State Network
Load forecasting
Neural Network
File in questo prodotto:
File Dimensione Formato  
PID6464377.pdf

accesso aperto

: Pre-Print (o Pre-Refereeing)
Dimensione 485.06 kB
Formato Adobe PDF
485.06 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1171227
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 1
social impact