Title :
A methodology for real time fault detection, isolation and correction
Author :
Pensick, Ellen C. ; Mulholland, John E.
Author_Institution :
Gen. Electr. Co., King of Prussia, PA, USA
fDate :
7/1/1993 12:00:00 AM
Abstract :
Mission critical timelines and availability requirements of large, complex ground stations demand some form of a distributed real-time system with an embedded rule-based expert advisor for recognizing and responding to changes within the system. The failure management optimization support system (FMOSS), which develops unique sets of status values, is discussed. The live status returned from the segment hardware forms a set of values, or signature vector, for each equipment string. The signature vector is compared to the signature that occurs when the equipment string is operational. If the processed status signatures do not match the signature that indicates the equipment string is operational, the expert advisor notifies the command center that a fault has been detected. The expert advisor continues processing until it has found a match from the pool of preengineered failure-mode signature vectors. Based on the match, Boolean logic is used to perform fault isolation and correction, with results provided to the command center.<>
Keywords :
Boolean functions; aerospace computing; diagnostic expert systems; distributed processing; failure analysis; pattern recognition; real-time systems; Boolean logic; FMOSS; availability requirements; distributed real-time system; embedded rule-based expert advisor; failure management optimization support system; failure-mode signature vectors; fault correction; fault isolation; ground stations; mission critical timelines; real time fault detection; segment hardware; Boolean functions; Control systems; Downlink; Fault detection; Ground support; Hardware; Monitoring; Real time systems; Satellite ground stations; Signal processing;
Journal_Title :
Aerospace and Electronic Systems Magazine, IEEE