Optimal Filter Design for Gear Fault Detection under Time-Varying Speed Conditions using Generalised Envelope Spectrum-Based Signal-to-Noise Objectives

Centrala begrepp

A generalised envelope spectrum-based signal-to-noise objective is proposed to derive optimal filter coefficients that enhance gear fault signatures and attenuate extraneous components under time-varying speed conditions.

Sammanfattning

The content presents a method for optimal filter design to enhance gear fault signatures in vibration signals under time-varying speed conditions. The key points are:

A Generalised Envelope Spectrum-based Signal-to-Noise (GES2N) objective is proposed that can be used to estimate existing objectives like the Indicator of Second-order Cyclostationarity (ICS2) and the Maximum Squared Envelope Spectrum Harmonic-to-Interference Ratio Deconvolution (MSESHIRD), as well as develop new envelope spectrum-based signal-to-noise objectives.

The GES2N objective is specifically derived for time-varying speed conditions, using the Velocity Synchronous Discrete Fourier Transform to calculate the squared envelope spectrum.

Five objective functions are derived from the GES2N formulation and compared against existing methods like the Cyclic Blind Deconvolution (CYCBD) and Adaptive CYCBD on three experimental gear fault datasets under time-varying speed.

The GES2N-Max-Np and GES2N-Mean-Np objectives are found to outperform the other methods in enhancing the gear fault signatures and attenuating extraneous components across the three datasets.

Statistik

The gear damage manifests at a cyclic order of 1.0 and its harmonics.
An extraneous impulsive component is present at a cyclic order of 5.72 and its harmonics.

Citat

"Fault detection primarily uses the fault component's prominence in the squared envelope spectrum, quantified by a squared envelope spectrum-based signal-to-noise ratio."
"New optimal filter objective functions are derived from the proposed generalised envelope spectrum-based signal-to-noise objective for machines operating under variable speed conditions."

Viktiga insikter från

Generalised envelope spectrum-based signal-to-noise objectives: Formulation, optimisation and application for gear fault detection under time-varying speed conditions

by Stephan Schm... på arxiv.org 05-03-2024

https://arxiv.org/pdf/2405.00727.pdf

Djupare frågor

How can the proposed GES2N objective be extended to handle multiple fault signatures simultaneously?

The proposed GES2N objective can be extended to handle multiple fault signatures simultaneously by incorporating multiple targeted cyclic order bands in the numerator and denominator of the objective function. Each targeted cyclic order band can correspond to a specific fault signature, allowing the objective to focus on enhancing the diagnostic information related to each fault signature. By adjusting the weighting matrices and weight vectors for each targeted band, the GES2N objective can effectively capture and enhance multiple fault signatures present in the vibration signals. This approach enables the optimal filter design to be tailored towards detecting and isolating various fault signatures in the system simultaneously.

What are the limitations of the envelope spectrum-based approach, and how can it be combined with other signal processing techniques for more robust fault detection?

The envelope spectrum-based approach has limitations in terms of its sensitivity to extraneous impulsive components, difficulty in distinguishing between different fault signatures, and challenges in handling time-varying operational conditions. To overcome these limitations and enhance fault detection robustness, the envelope spectrum-based approach can be combined with other signal processing techniques such as:

Time-Frequency Analysis: Utilizing time-frequency analysis techniques like wavelet transform or spectrogram analysis can provide additional insights into the signal characteristics and help in identifying specific fault signatures in different time-frequency domains.

Machine Learning Algorithms: Integrating machine learning algorithms such as neural networks or support vector machines can improve fault classification and pattern recognition by learning from the complex patterns present in the vibration signals.

Statistical Analysis: Incorporating statistical methods like hypothesis testing or anomaly detection can help in identifying abnormal behavior in the vibration signals and distinguishing between fault signatures and noise components.

Adaptive Filtering: Implementing adaptive filtering techniques like Kalman filtering or recursive least squares can adapt to changing operational conditions and enhance the fault detection performance under varying speed conditions.

By combining the envelope spectrum-based approach with these complementary signal processing techniques, a more comprehensive and robust fault detection system can be developed, capable of effectively detecting and diagnosing multiple fault signatures in rotating machinery.

Can the GES2N objective be adapted for online, real-time optimal filter design for gear fault detection under time-varying conditions?

Yes, the GES2N objective can be adapted for online, real-time optimal filter design for gear fault detection under time-varying conditions by implementing an adaptive filtering approach. By continuously updating the filter coefficients based on the real-time vibration signals and adjusting the objective function parameters to account for changing operational conditions, the GES2N objective can be optimized dynamically to enhance fault detection performance in a time-varying environment.
Key steps for adapting the GES2N objective for online, real-time optimal filter design include:

Continuous Data Acquisition: Implement a real-time data acquisition system to capture vibration signals from the rotating machinery.
Adaptive Filter Update: Develop algorithms to update the filter coefficients based on the incoming data and the evolving fault signatures.
Dynamic Objective Function: Modify the GES2N objective to incorporate real-time feedback and adjust the weighting matrices and parameters to adapt to changing fault conditions.
Feedback Mechanism: Implement a feedback mechanism to evaluate the filter performance and adjust the objective function parameters accordingly.
Integration with Control Systems: Integrate the real-time optimal filter design system with the machinery's control systems for seamless operation and fault detection.

By integrating these components and ensuring real-time adaptability, the GES2N objective can be effectively utilized for online, real-time optimal filter design for gear fault detection under time-varying conditions, enhancing the system's fault detection capabilities and reliability.