Model Predictive Control of pH Neutralization Process based on Manifold Gaussian Process

pH neutralization process is characterized with strong non-linearity and time-delay. Model Predictive Control is a common method to handle the control problem of such industrial processes. The main difficulty of the Model Predictive Control is to build a suitable predictive model that represents the inherent features of considered plant. In this paper, Manifold Gaussian Process is used to model the pH neutralization process, based on which the Model Predictive Control of system is realized. The first simulation results prove that excellent nonlinear modelling ability of Manifold Gaussian Process, and the following experiments are the further proof of which Model Predictive Control integrated with Manifold Gaussian Process has better robustness, stability and rapidity.