2005 · 25 citations · 10 references
EngineeringSystems EngineeringModel-based Systems EngineeringPower ElectronicsUncertainty QuantificationDrive SystemsElectric Motor Drive SystemsElectric Motor DrivesFault DetectionFault DiagnosisSystem IdentificationElectric Motor Fault DiagnosisFault Diagnosis For Nonlinear SystemsFault Diagnosis In Motor DrivesMachine Fault DiagnosisModel-based DiagnosisModel-based Fault DiagnosisNonlinearityRobustness
The performance of model-based fault diagnosis (FD) algorithms is degraded by modeling errors caused by process uncertainties. FD robustness against process uncertainties is discussed in this paper. In the proposed FD schematics, robustness is improved by using adaptive residual generators, in which the advanced on-line identification plays a key role. The approach is investigated on a permanent magnet synchronous motor drive system with inherent nonlinearity and time-varying parameters caused by process uncertainties. Stator resistance and q-axis inductance are considered as possible sources of system uncertainties under different operating conditions. Mathematical decoupling simplifies both identification and fault detection designs for the drive monitoring process. Preliminary simulation results are provided.
10
New Results in Linear Filtering and Prediction Theory
R. E. Kalman, R. S. Bucy · Journal of Basic Engineering · 1961 · 6.3K citations
Pétros Ioannou, Jing Sun · 1995 · 5.7K citations