Nonstationary modelling approaches of neutralization process for model-based control

In this paper, two simplified models of the neutralization process are investigated in terms of the modelling accuracy. Both models have relatively simple and affine form, which is the key requirement for their application as a basis for the model-based control. One of these models is a simple input-output model while the second one has a phenomenological origin. Both of them include the estimation of the unknown time-varying parameter representing modelling inaccuracies. The estimation procedure is based on the scalar form of the Recursive Least-Squares method. The modelling accuracy of both models is tested by simulation in the application to the complex model of the neutralization process and on the basis of the real measurement data from the laboratory neutralization pilot plant.