Diagnostic Classifier Ensembles: Enforcing Diversity for Reliability in the Combination

The construction of reliable machinery fault diagnostic systems is investigated in this thesis. The idea of using unitary sensorial information to develop automated fault classifiers is studied. The domain of fault diagnosis is used to investigate the concept of enforcing methodological diversity in the solutions with a view to obtaining robust and reliable systems by combining the decisions of several independent classifiers. A twin cylinder, high speed, 4-stroke diesel engine is used as an exemplar of the class of mechanical machines. Several commonly occurring faults symptomatic of early stages of fault development are induced in the engine. Data in the form of cylinder pressure and vibration are acquired. Orthogonal wavelet transforms, principal component analysis and domain expertise of the engine cycle are used to extract features from the vibration and pressure signals. Several artificial neural net classifiers are trained with these features, after establishing that simpler visualisation techniques and derived parameters do not provide cues in discriminating between the classes in the data. Statistical models are used to evaluate the diversity within the methodologies used to create the classifiers. These evaluations are used to propose the most effective assemblage of majority voting diagnostic systems. ‘Sensorial’ diversity is suggested as an extension to the concept of methodological diversity and it is shown that the most robust systems consist of solutions from the most diverse methodologies. A novel metric for diversity is suggested which is shown to have a definite positive correlation with the classification accuracy as well as with the robustness of the ensembles of diagnostic systems.