ualitative Multi iagnosis of Continuous ehavioral Modes

Most model-based diagnosis systems, such as GDE and Sherlock, have concerned discrete, static systems such as logic circuits and use simple constraint propagation to detect inconsistencies. However, sophisticated systems such as QSIM and QPE h ave been developed for qualitative modeling and simulation of continuous dynamic systems. We present an integration of these two lines of research as implemented in a system called QDOCS for multiple-fault diagnosis of continuous dynamic systems using QSIM models. The main contributions of the algorithm include a method for propagating dependencies while solving a general constraint satisfaction problem and a method for verifying the consistency of a behavior with a model across time. Through systematic experiments on two realistic engineering systems, we demonstrate that QDOCS demonstrates a better balance of generality, accuracy, and efficiency than competing methods. Kuipers 1984). Our work uses QSIM (Kuipers 1994) as the modelling language and applies a very general diagnostic technique to models described in this language. Previous approaches to diagnosing faults in systems described with QSIM models have been limited in scope and have been unable to work with fault modes (Ng 1990; Lackinger & Nejdl 1991) or have made a singlefault assumption (Dvorak 1992). Most previous work on model-based diagnosis (Reiter 1987; de Kleer & Williams 1987) h as concentrated on static systems and is generally insufficient to diagnose continuous dynamic systems. Few of the other approaches to diagnosis of continuous systems (Oyeleye, Finch, & Kramer 1990; Dague et al. 1991) h ave made use of a general modelling language such as that provided by QSIM or used any of the general diagnostic formalisms introduced by Reiter or DeKleer.