基于MLD模型的混杂预测控制研究-控制科学与工程专业论文.docxVIP

Research Research of Hybrid Predictive Control Based on MLD Model Abstract For modem industrial process control，on the one hand high-precision and lligh—qulity produet control requirements are fulfilled under the circumstance that multivafiable，nonlinearility and uncertainty features exist in process，on the other hand combinational optimal control problems of continuous characteristics，discrete events and if-then-else rules are a part of industrial process．Therefore，it is more apparent that hybrid features are more and more important in modem industry,and the complexity of computation has also been increasing．To consider these conditions，this paper makes use of MLD(Mixed Logical Dynamical)model to study the modeling and predictive control problems of hybrid system，main research works and achievements are summarized as following： Logical methemeties knowledge which will be needed for MLD model，and methods which can be applied into modeling have been discussed in detail，in addition， the procedure and methods to establish MLD model are elaborated on through the MLD modeling of a typical spherical three—tank system． Optinization control algorithm of MLD model is studied in detail，which contains the solution problems of optimal control and predictive contr01．The inproved algorithm of optimal control for MLD model has been acquired and the simulation of improved one has been carried out．Hybrid predictive control simulation based on MLD model is carried out through spherical three—tank system． The reasons of infeasibility of constrained predictive control are analyzed，then hybrid approach is proposed for the problem of constraints prioritization and feseability,simulation results of Wood-Berry and Shell column authenticate the effectiveness of this method，at the same time the improved algorithm for hybrid one is proposed and the simulation is carried out． PFC(Predictive Functional Contr01)approach is introduced，the basic theories are discussed discussed in detail，and the