loading page

Benchmarking Reservoir Computing for Residential Energy Demand Forecasting
  • +4
  • Karoline Brucke,
  • Simon Schmitz,
  • Daniel Köglmayr,
  • Sebastian Baur,
  • Christoph Räth,
  • Esmail Ansari,
  • Peter Klement
Karoline Brucke
DLR-Institute of Networked Energy Systems

Corresponding Author:[email protected]

Author Profile
Simon Schmitz
DLR-Institute of Software Technology
Daniel Köglmayr
DLR-Institute for AI Safety and Security
Sebastian Baur
DLR-Institute for AI Safety and Security
Christoph Räth
DLR-Institute for AI Safety and Security
Esmail Ansari
Fraunhofer Institute for Manufacturing Technology and Advanced Materials IFAM
Peter Klement
DLR-Institute of Networked Energy Systems


In the energy sector, accurate demand forecasts are vital but often limited by the available computational power. Reservoir computing (RC) or echo-state networks excel in chaotic time series prediction, with lower computational requirements compared to other recurrent network based methods like LSTMs. Next-generation reservoir computing (NG-RC) is a newer, more efficient variant of classical RC originating from nonlinear vector autoregression and therefore missing the randomness of classical RC. In our study, we evaluate RC and NG-RC for day-ahead energy demand predictions on four data sets and compare it to LSTMs and a naive persistence approach. We find that NG-RC outperforms all other methods when considering the root mean squared error on all data sets but struggles with very small or zero demands. Additionally, it offers a very computationally effective hyperparameter optimization and excels in replicating the inherent volatility and the erratic behavior of energy demands.
07 Feb 2024Submitted to TechRxiv
12 Feb 2024Published in TechRxiv