A New Approach to Model-Based Diagnosis Using Probabilistic Logic

We describe a new approach to model construction using transfer function diagrams that are consequently mapped into generalized loopy logic, a first-order, Turing-complete stochastic language. Transfer function diagrams support representation of dynamic systems with interconnected components. We demonstrate how these diagrams provide interfaces to a context-sensitive probabilistic modeling system (COSMOS). As a result, interfaces as well as the notion of context underlying COSMOS are successfully used for model-based diagnosis. This paper describes transfer function diagrams and how they are incorporated into COSMOS. We illustrate our approach with a practical example taken from a “pump system” modeling problem.