Study On Multivariable Predictive Functional Control And Its Applications

Model Predictive Control (MPC) is a model of successful applications of modern control theory in industrial process control and Predictive Functional Control (PFC) is one of the newest research field of MPC. The studies and applications of PFC are discussed and the Multivariable Predictive Functional Control (MPFC) algorithm for first-order plus time delay is proposed in this thesis. The simulations of MPFC algorithm for two practical plants are also presented. The main content of the present research is as follows:1. An overview of MPC, PFC and study of multivariable systems is presented.2. The principle of PFC is introduced from three aspects: predictive model, error compensation and optimal algorithm. The basic PFC algorithm for first-order plus time delay system is developed. The closed-loop performance analysis for PFC control system is studied on the basis of Internal Model Control (IMC) method.3. Classification of the multivariable system model is provided. Considering a large class of Multi-Input and Multi-Output (MIMO) system that can be equivalently described as first-order plus time delay plant in each process channel, Multivariable PFC (MPFC) algorithm is presented. Computer simulation results demonstrate the effectiveness and robustness of this algorithm.4. The MPFC algorithm is applied to two multivariable industrial processes, including distillation control and basis weight & moisture control in paper-making process. The simulation results show the satisfying performance, the strong robustness and disturbance rejection of the strategy. And it is quite practical for this algorithm to be put into industrial control considering of its simplicity.The thesis concludes with a summary and perspectives of future research of PFC.