Structuring diagnostic knowledge for large-scale process systems

A set of guidelines is described for generating an initial organization of knowledge for distributed diagnosis of a process plant. The diagnostic know ledge is organized hierarchically by primary processing systems (commonly f eed, reaction, and separation in chemical plants), subsystems, components, behaviors and malfunction modes. The resulting classification hierarchy dec omposes the diagnostic problem solving into coordinated, distributed module s, where different modules may use different methods to address specific lo cal subproblems. Classification hierarchies, organized in this way, provide effective modularity for organizing large-scale, knowledge-based diagnosti c systems, which are difficult to construct without pertinent organizing pr inciples. Such hierarchies provide a framework for systematic knowledge acq uisition and maintenance. Application of the guidelines emphasizes readily available sources of knowledge, considers common design and operating objec tives of process plants, draws upon operating expertise and builds on gener ic process characteristics. Application is illustrated for a fluidized cata lytic cracking unit and a paraxylene production unit. (C) 1998 Elsevier Sci ence Ltd. All rights reserved.