Particle swarm optimization based neural-network model for hydro power plant dynamics

This paper addresses the modeling of hydro power plant dynamics using neural network approach. The cost function as root mean square error is optimized by particle swarm optimization technique. The identification performance is compared with fuzzy models based on GK clustering algorithm in application to study hydro power plant dynamics. It is found that the response obtained from the NN model is comparable to those determined by fuzzy model with much significance to nature of input-output variables used for modeling.