A comparative study of water wall model with a linear model and a neural network model

Abstract A water wall system is one of the most important systems used in boilers of thermal power plants. In this study, we assume that an exact nonlinear water wall model is given. Then, from the viewpoint of controller design, we compare the effectiveness of two models, a linear model and a neural network model. The linear model is developed by linearizing a nonlinear model at an operating point. The neural network model is developed using simulation data from a nonlinear model. To compare the qualities of these two models, a PRBNS (Pseudo Random Binary Signal) is applied to each model, and the errors of the models are compared. In comparison to the nonlinear model, two models performed adequately. The neural network model achieved a better performance than the linear model.