The paper presents the concept of using a Blind Source Separation (BSS) algorithms to diagnose localized faults in rolling element bearings. These bearing faults usually result in strong harmonics of the fault frequencies along with sidebands in the spectrum of the vibration signals. We have shown that separation of vibration signal components containing these fault frequencies enables more effective bearing diagnostics. We have used linear BSS algorithms as well as nonlinear version. The conducted experiment showed an increase in the effectiveness of fault detection for all types of simulated bearing faults. The results achieved may enable the construction of an expert system in the future that diagnoses the condition of technical devices based on the analysis of the vibro- acoustic signal generated by the operating device.
Application of Blind Source Separation Methods in the Diagnosis of Rolling Bearings
Alessandro Ruggiero
2025
Abstract
The paper presents the concept of using a Blind Source Separation (BSS) algorithms to diagnose localized faults in rolling element bearings. These bearing faults usually result in strong harmonics of the fault frequencies along with sidebands in the spectrum of the vibration signals. We have shown that separation of vibration signal components containing these fault frequencies enables more effective bearing diagnostics. We have used linear BSS algorithms as well as nonlinear version. The conducted experiment showed an increase in the effectiveness of fault detection for all types of simulated bearing faults. The results achieved may enable the construction of an expert system in the future that diagnoses the condition of technical devices based on the analysis of the vibro- acoustic signal generated by the operating device.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.