Development of a Nonlinear Least Squares Method for the Model Parameterization of a SCR Model

In model based selective catalytic reduction (SCR) control, because of the nonlinearities and immeasurable properties of the SCR physical background, model parameterization is always a problem to be solved. In this paper, a new method based on the nonlinear least square method is developed to estimate the parameters in a SCR model. The SCR system state functions are specially designed in terms of the parameter characteristics of the SCR model. To simplify the mathematical process, a nonlinear least square MATLAB function is used in practice. The regression parameters are used in a SCR simulation model; the simulation results show great consistency with the experimental results.