Jixie chuandong (Jan 2024)
Modulation Signal Bispectrum Slice Overall Average Feature Extraction of Low-speed Reciprocating Bearing Faults
Abstract
Yaw bearings and slewing bearings in wind turbine systems, rotating support bearings in aerospace launch towers, and slew bearings in cranes and excavators all exhibit characteristics of low-speed reciprocating motion. Diagnosing faults in low-speed reciprocating motion bearings presents significant challenges: the impact forces from damage contact are minimal at low speeds, resulting in weak damage impact signals. Furthermore, the interference from deceleration and directional change impacts is substantial, and long signals covering multiple reciprocating cycles lack periodicity. To address these challenges, a method for diagnosing faults in low-speed reciprocating motion bearings is proposed. This method is based on modulation signal bispectrum （MSB） slice-wise overall averaging. The approach begins with signal resampling, which involves tracking zero-crossings in the rotational speed to resample vibration signals. Subsequently, it separates short signals from the resampled data based on encoder signals, forming a collection of short signals corresponding to individual reciprocating cycles. Each short signal undergoes MSB analysis to generate the MSB bicoherence slice spectrum. The optimal carrier frequency and its associated modulation signal slice spectrum are determined from the bicoherence slice spectrum. Finally, an overall average of the MSB modulation signal slice spectra from the short signal collection results in the overall averaging feature. Validation of the method using fault test data demonstrates its effectiveness in diagnosing faults in low-speed reciprocating motion bearings.
