Adaptive combined HOEO based fault feature extraction method for rolling element bearing under variable speed condition

Abstract

It is still a challenge for condition monitoring and fault diagnosis for rolling element bearings working under variable speed, while some conventional diagnostic methods are useless. Order tracking is commonly used as an effective tool of the non-stationary vibration analysis for rotating machinery and tacho-less order tracking may be more applicable for practical situations, while the key point is to obtain more accurate instantaneous rotating speed. What’s more, the collected bearing fault vibration signals always contain strong background noise that greatly affects the result of fault feature extraction. To solve these problems, a fault feature extraction method is proposed in this study. The Chirplet-based approach was used to estimate some obvious harmonics of instantaneous rotating speed and the average value of these components was regarded as the final instantaneous rotating speed to reduce the estimation error. Since the higher order energy operator (HOEO) can not only improve the signal-to-noise ratio and signal-to-interference ratio, but is also easily applied, an adaptive combined HOEO method based on hybrid particle swarm optimizer with sine cosine acceleration coefficients (H-PSO-SCAC) was constructed to enhance the impulse components, and then, the fault features were extracted by the order spectrum analysis. Simulation and experimental results indicate that the proposed algorithm is effective for rolling element bearings’ fault diagnosis under variable speed condition.