ComputingComputer ScienceSystems EngineeringAlgorithmsDiscrete MathematicsTime Series AnalysisDiagnosisFault DetectionFault DiagnosisFault ModelingExperimental StudiesFault Detection In Control SystemsFault EstimationRecursive Algorithm DesignSliding Window AlgorithmsSliding Window AnalysisSystem Fault Diagnosis
The development of a fault detection scheme for linear jump discrete systems is addressed. The system's parameter vector is identified through a sliding-window orthotopic set membership identification scheme. A recursive algorithm predicts the interval of the system's output based on the identified parameter bounds. The fault is detected when the actual system output is not within the predicted 'worst case' interval output, or when the actual parameter vector is not within the identified interval. Simulation and experimental studies are offered to illustrate the scheme's performance.
14
A survey of design methods for failure detection in dynamic systems
Alan S. Willsky · Automatica · 1976 · 2.1K citations
Stochastic stability properties of jump linear systems
Xiangbo Feng, Kenneth A. Loparo, Y. Ji et al. · IEEE Transactions on Automatic Control · 1992 · 724 citations