Chemical processing systems are typically nonlinear and complex. Control operators in industry tend to rely on their own experience and heuristics (or simple common sense) to control a complicated process. Fuzzy logic control is an efficient method which can easily incorporate heuristics. Fuzzy logic controls are based on fuzzy set theory which is designed to allow quantitative measurements of qualitative descriptions such as tall, large, small, etc. used frequently in human reasoning. A fuzzy set is a set of numbers having membership values between 0 and 1, where 0 signifies weakest membership and 1 signifies strong membership. This research project will investigate pseudo-fuzzy logic controllers (PFLC) for the control of complex nonlinear chemical processes. PFLC is based on conventional fuzzy logic controllers but it translates heuristics into algebraic equations (fuzzy logic sets use decision tables) with tunable parameters. The methodology includes the derivation of pseudo-fuzzy logic models which incorporate confluence methods for obtaining model structure and modulating function methods for system identification. Heuristics are then translated to PFLCs. System performance can be analyzed using modern control methods such as frequency response, Lyapunov or H infinity methods. New rules can be obtained by explicitly solving for the control variables from pseudo-fuzzy logic models. To investigate the practical applications of PFLC, an experimental study of temperature control of a polymerization reactor will also be performed. The system will be the polymerization of polydimethylsiloxanes.
