Artificial immune system via Euclidean Distance Minimization for anomaly detection in bearings

摘要: Abstract In recent years new diagnostics methodologies have emerged, with particular interest into machinery operating in non-stationary conditions. fact continuous speed changes and variable loads make non-trivial the spectrum analysis. A means a characteristic fault frequency related to damage that is no more recognizable spectrum. To overcome this problem scientific community proposed different approaches listed two main categories: model-based expert systems. context paper aims present simple system derived from mechanisms of immune called Euclidean Distance Minimization, its application real case bearing faults recognition. The method simplification original process, adapted by class Artificial Immune Systems, which proved be useful promising fields. Comparative results are provided, complete explanation algorithm functioning aspects.