A hybrid model for SSME diagnosis

The authors present the knowledge-based system EDIS (Engine Data Interpretation System) which incorporates a novel approach to the coordination and cooperation of diverse knowledge sources in the task of Space Shuttle Main Engine (SSME) diagnosis. EDIS achieves tight coupling between heterogeneous paradigms and procedures, such as heuristic and model-based reasoning. Various reasoning methods share data hypotheses and intermediate conclusions through a blackboard without the need for translation. Each diagnostic process is subdivided into small steps so that steps from different methods can be selected and executed in arbitrary order by the diagnostic strategy. A prototype of EDIS has been implemented which demonstrates how the capabilities of very different diagnostic paradigms can be combined to form an integrated hybrid diagnostic paradigm