The Nonlinear Model Predictive Control And Its Applications In The Polyester Industry

Predictive control is a class of digital control algorithms. The population of these methods in industry is due to the fact that they don't require precise model prediction, model can be obtained by simple experiment; the uncertainty is compensated for model mismatch or disturbance existed by using receding horizontal optimization; and the algorithm can be easily extended to constrained and large delay processes and deal with constrained multivariable problems efficiently.With the development of modern industry and progress of science technology, the process industry has become more and more complex and characterized with strong nonlinearity. Therefore, linear model predictive control (LMPC) may no longer satisfactory. Predictive control based on nonlinear model (NMPC) has become an research issue of great importance in the control engineering fields.Some problems of predictive control, especially nonlinear predictive control are researched in this dissertation, and the main research works are as follow:(1) The basic concepts and developments of predictive control are introduced with the focus on nonlinear predictive control, and some simulations are given as well.(2) A predictive control method with nonlinear state space model is proposed on the bases of much study of the nonlinear predictive control, large amount of simulations are given as well.(3) The nonlinear predictive control algorithm has been applied to the esterification process of the polythylene terephthalate(PET) industry. And the optimized control over the quality indexes of the esterification products has been realized. The result is quite satisfactory throughout the actual operation.