Model-based Vs. Model-less AI Fault Diagnosis
The main aim of diagnostics is the identification of cause-effect chain that is not fully achieved by reliance only on available machine parameters.
Restriction of the analysis to deterministic systems means that modelling of faulty machines can be done successfully. Relying on these models, the behaviour of a specific machine can be accurately analysed but the extension of the results to other machines is not straightforward. Another issue is the accuracy of the model and if it can apply to all modes of the machine. The majority of machine models only correspond to the linear part of the system. For example in overload of the motor where the motor is in non-linear performing area or in short winding faults where the number of windings and eventually the core is changing the linear model is not applicable.
Artificial Intelligence (AI) on the other hand is not as accurate in the linear part of the system but can extend itself to non-linear part of the operation or even other similar machinery of it’s type.