Extending conventional loop design techniques to fuzzy logic control

Fuzzy Logic Control may be described as a knowledge-based linguistic control strategy. The knowledge base is represented by a set of control rules. These rules allow human intuition and experience to become the control strategy. In a conventional control system the control law is an analytic function of the process state. In fuzzy logic control systems the control law is composed of if condition-then action type linguistic statements in which the action, or consequent, is a function of the condition, or antecedent.; Since its inception in 1965, fuzzy logic control has found application in the process industries. In spite of any number of successful applications, it has not been widely accepted by the control engineering community. This may be attributed to the lack of a sound foundation for analysis and design in traditional engineering terms. The objective of this dissertation is to offer such a foundation.; This objective is accomplished in three distinct ways. First, a mathematical analysis of the current techniques used in the realization of a fuzzy logic controller is performed. This analysis defines how the various fuzzy logic controller parameters influence its dynamic properties. This is the first step in the process of establishing fuzzy logic control in traditional engineering terms. This analysis defines the relationship between the fuzzy logic controller parameters and common engineering performance measure.; Secondly, alternatives to current techniques are proposed. These alternatives transform the nonlinear controller into a linear model. This model is expressed in traditional engineering terms and can be used with any of the conventional analysis and design strategies.; Finally, a design strategy is proposed for both known and unknown processes. This strategy, based on the proposed alternatives, utilizes the linear model to achieve expected performance standards. The concept of trajectory tuning is introduced as a means of improving performance by nonlinearizing the response trajectory. A case study is included to illustrate how superior performance can be achieved.