Real-time Intelligent Control Of CSTR Plant

By means of AI & Pattern Recognition techniques, and with the help of Modern Control Theory & Fuzzy Sets, the real-time intelligent control of a Continuous-Stirred Tank Reactor, which is widely used in the polymerization industry, is studied in this thesis .This thesis consists of six parts .Chapter 1 introduces the objective and target of research , as well as the state-of-art of the CSTR control and Ai development. the contents of this thesis is also briefly described .Chapter 2 briefly introduces the trend of system modelling and the mathematical model of CSTR system . The real-time state estimation strategy which is based on Extended Kalman filtering is also presented .Chapter 3 is devoted to pattern classification . The approach for the training of pattern classifiers is carefully studied , and an improved Perceptron Approach is designed for the construction of Expert Controller .Chapter 4 offers a Rule-Based Expert Controller for the real-time intelligent control of CSTR plant .