Systems modeling and simulation methodologies are useful for the analysis and design of industrial systems. Special purpose simulation languages have been developed to simplify the task of creating simulation models. While these languages make the programming portion of model creation more efficient, computer programming skills remain a requirement for effective modeling, and more importantly, for model comprehension by decision makers who are likely to be non-programmers. The logic programming paradigm, in which programs consist of facts and rules expressed in a near natural language format, can make simulation models more comprehensible to decision makers. Complex systems such as an industrial enterprise operating in physical, biological, and socioeconomic environments generally have many ill defined, or imprecisely known, parameters that are typically approximated by exact numbers or represented with random variables. Fuzzy set theory is a promising alternative to probability theory for explicitly treating these aspects of a system.;The objective of this research was to extend the utility of simulation modeling and analysis by modeling aspects of complex industrial systems that cannot be adequately addressed with conventional methodologies. To achieve this, a fuzzy set theoretic formalism is developed for knowledge-based modeling and simulation of industrial systems. This formalism, PRISM (Package for Intelligent Simulation in Mathematica), is implemented as a prototype knowledge-based simulation environment which was developed in Mathematica.;Models of industrial processes were implemented based upon fuzzy set theoretic specification and fuzzy inferencing. These fuzzy models capture complex aspects of the target systems, such as environmental risk perception and system fault diagnosis, that are difficult to capture with conventional methods. Research questions addressed are the requirements for special purpose fuzzy logic programming systems for simulation and modeling, problem domain representation, model verification and validation, and user interaction with simulation models based upon fuzzy logic.
展开▼
