An application of qualitative reasoning to process diagnosis: automatic rule generation by qualitative simulation

AQUA, a knowledge acquisition tool for the process diagnosis domain, is presented. AQUA uses a qualitative model in guiding the construction of a diagnostic rule base. Since online real-time diagnosis of the real process plant requires a huge rule base, a systematic way of building the rule base is proposed by a qualitative version of fault simulation on the qualitative model. The qualitative model is modified interactively, using qualitative simulation and qualitative diagnosis. To further enhance the knowledge acquisition process, the knowledge from which the model is constructed is divided into two types: plant-common knowledge and plant-specific knowledge.<<ETX>>