Fault Detection Frameworks
Fault Detection Frameworks
benbouzidSignalProcessingFault2020
Reference: (benbouzidSignalProcessingFault2020, link, DOI)
Model-based approaches for diagnosis. Model-based method: It is used to measure the deviation between the model output and the actual machine output and then predict a potential failure signature. The goal is to generate several symptoms indicating the difference between nominal behavior and abnormal operating conditions.
Signal-based approaches for diagnosis. Any kind of fault modifies the symmetrical properties of electrical machines. Therefore, characteristic fault frequencies appear in some physical signals issued from sensors. The analysis of these signals allows to enhance the knowledge about a specific fault, its impact on intrinsic parameters of the machine, and its frequency signature. Signal analysis is performed using suitable signal conditioning and processing techniques for fault features extraction. Then, a fault decision algorithm performed for distinguishing faulty cases from healthy ones and classification purposes.
niuDataDrivenTechnologyEngineering2017
Model-based and data-driven fault diagnosis. a Model-based, b Data-driven