Short-Time Variance Providing Evidential Reference Frequency for Lock-in Amplifier in Fault Diagnosis of Rolling Bearings

Purpose: The purpose of this study is to propose a framework employing Lock-in amplifier for fault diagnosis of rolling bearings using vibration signals, addressing the challenges posed by noise interference and inaccurate estimation of Lock-in amplifier’s reference frequency. Methods: By studying the properties of the pulse chains triggered by faults, the proposed framework utilizes the Short-time Variance of vibration signals to determine the bearing fault characteristic frequency, providing a reliable reference frequency for the Lock-in amplifier. This approach aims to overcome the limitations of previous methods reliant on geometric relationships for fault frequency estimation. Results: It has been proved that Short-time Variance can effectively identify sparse pulses in vibration signals. The Short-time Variance spectrum effectively characterizes bearing fault information and outperforms the benchmark method envelope spectrum in high-noise environments. Validation of the proposed framework is conducted using simulated signals and experimental data sourced from a public dataset. The results demonstrate the effectiveness of the framework in accurately diagnosing bearing faults through vibration signals. Conclusion: This study confirms that directly applying theoretical fault characteristic frequencies as the reference frequency for a Lock-in amplifier can have a significant negative impact on diagnostic performance. It has been concluded that the proposed framework offers a robust approach to fault diagnosis of rolling bearings. By leveraging Lock-in amplifiers and the Short-time Variance, it provides accurate fault detection, thus ensuring the safe and reliable operation of mechanical equipment.