A Model-Based Failure Detection, Isolation and Recovery System

Abstract : The Failure Detection, Isolation and Recovery (FDIR) in the International Space Station Alpha (ISSA) requires timely monitoring and diagnosis of failures so that recovery actions can be employed to safeguard the mission and the life of crew. Using traditional methods for representation of domain knowledge and for diagnosis proves to be ineffectual because of the scale, complexity and dynamics of ISSA. Model-based approach for representing systems and for diagnosis is an attractive and feasible solution. We have developed and field tested a model-based real-time robust monitoring and diagnostic system for ISSA and other aerospace systems. The system is represented using hierarchical and multiple- aspect models, which include representation of functional structure as well as the physical component assemblies. A discretized model of the failures and their effects is represented using timed failure propagation graphs. The monitoring mechanism is modeled by using a discretized sensor space, with mechanisms for sensor validation. The diagnostic reasoning applies structural and temporal constraints for the generation and validation of fault hypotheses using the predictor-corrector principle. The diagnosis is generated in real-time amid an evolving alarm scenario, and uses progressive deepening control strategy. The robust diagnostic system has been tested and demonstrated using ISSA models obtained from the Boeing Company.