Estimación de modelos lineales para el control predictivo de pequeñas centrales hidroeléctricas

Abstract

This paper presents the modeling of a Small Hydroelectric Plant (PCH), by obtaining linear prediction models. The mathematical model of the system, provides all your hydraulic system, which runs from the reservoir tank, through the tunnel for the transport of water resources, load lines, up to the generation units. Subsequently, the system proceeds to modeling, using a set of prediction models like ARX models, ARMA X, Box-Jenkins and Output Error (OE). The results obtained for each of the above models were evaluated versus nonlinear dynamic model that represents the dynamic behavior of the system. This evaluation was based on error analysis and according to the performance index of each model. Thus, the best results were obtained by the OE model as this shows a direct way as the error may affect the system output. Finally, the development of this model can be used in the implementation of predictive controllers for load distribution (power generation) in Small Hydroelectric Plant, based on the conditions of water availability in the reservoir tank.