Fault detection for planetary gearbox based on an enhanced average filter and modulation signal bispectrum analysis

Area of Science:

• Mechanical Engineering
• Vibration Analysis
• Signal Processing

Background:

• Planetary gearboxes generate transient impulses crucial for fault diagnosis.
• Vibration signals are often contaminated by noise and harmonics, obscuring these impulses.
• Existing methods like mathematical morphology (MM) struggle with bidirectional impulses.

Purpose of the Study:

• To develop a novel method for accurate planetary gearbox fault detection.
• To enhance the extraction of impulse features from noisy vibration signals.
• To improve the accuracy and reliability of machinery fault diagnosis.

Main Methods:

• Proposed an enhanced average (EAVG) filter, combining MM operators for bidirectional impulse extraction.
• Optimized EAVG filter performance using kurtosis for adaptive structural element length selection.
• Applied modulated signal bispectrum (MSB) to EAVG-filtered signals to extract fault-related modulation components.

Main Results:

• The EAVG filter effectively extracts both positive and negative impulses, outperforming traditional MM operators.
• EAVG-MSB successfully decomposed nonlinear modulation components, revealing fault characteristics.
• Validated against CMFH and ACDIF, EAVG-MSB demonstrated superior accuracy in detecting sun gear chipped tooth, misalignment, and bearing inner race faults.

Conclusions:

• The EAVG-MSB method offers a robust and accurate approach for planetary gearbox fault diagnosis.
• This technique effectively overcomes noise and harmonic interference in vibration signals.
• EAVG-MSB proves feasible for real-world condition monitoring and fault diagnosis applications.