24th EANN 2023, 14 - 17 June 2023, León, Spain

Adaptive model for industrial systems using Echo State Networks

José Ramón Rodríguez-Ossorio, Antonio Morán, Juan J. Fuertes, Miguel A. Prada, Ignacio Díaz-Blanco, Manuel Dominguez


  When a model of an industrial system is developed, it is expected that this model performs consistently when applied to other identically designed systems. However, different operating hours, degradation or maintenance, among other circumstances, cause a change in the dynamics of the system and result in the model not performing as expected. For this reason, it is necessary to build a model that continuously adapts to changes in the dynamics of the system, in order to handle such deviations and thus reduce the estimation error. This paper proposes the development of an adaptive model based on Echo State Networks to estimate the level of a water tank. For this purpose, two identically designed industrial pilot plants are used, taking one of them as a reference for the parameterization, training and validation of the model, and applying the developed model to the other one in order to evaluate the adaptation to changes in the dynamics of the system.  

*** Title, author list and abstract as seen in the Camera-Ready version of the paper that was provided to Conference Committee. Small changes that may have occurred during processing by Springer may not appear in this window.