Control of pH neutralization process using multiple models and predictive control

Abstract

The paper deals with the problem of modelling nonlinear processes using the Local Model Network (LMN). The idea is based on development of the local linear models for the whole operating range of the controlled process. The nonlinear plant is then approximated by a set of locally valid submodels, which are smoothly connected using the validity function. For saving the computational load, linear model is obtained by interpolating these linear models at each sample instant and then used in GPC framework to calculate the future behaviour of the process. It avoids the time consuming quadratic optimization calculation, which is normally necessary in nonlinear predictive control. The proposed approach is applied to a simulated nonlinear pH neutralization process.