A Hydraulic Pump Fault Diagnosis Method Based on the Modified Ensemble Empirical Mode Decomposition and Wavelet...

Area of Science:

• Mechanical Engineering
• Condition Monitoring
• Signal Processing

Background:

• Axial piston pumps are critical in many hydraulic systems.
• Diagnosing faults in these pumps is challenging due to signal complexity.
• Existing methods may lack accuracy or speed for effective fault detection.

Purpose of the Study:

• To develop an integrated, accurate, and fast fault diagnosis method for axial piston pumps.
• To address the limitations of current diagnostic techniques for hydraulic machinery.
• To improve the reliability and maintenance efficiency of axial piston pumps.

Main Methods:

• Utilizing modified ensemble empirical mode decomposition (MEEMD) to decompose complex vibration signals.
• Applying autoregressive (AR) spectrum analysis to extract fault characteristics from signal components.
• Employing a wavelet kernel extreme learning machine (WKELM) for high-accuracy fault classification.

Main Results:

• The proposed method achieved 100% recognition accuracy for single slipper wear, single slipper loosening, and center spring wear faults.
• The integrated diagnostic method demonstrated significantly faster fault diagnosis times (0.002 s) compared to traditional methods.
• Comparative analysis showed superior performance over BP neural network, SVM, and ELM methods.

Conclusions:

• The integrated MEEMD, AR spectrum energy, and WKELM method offers a highly accurate and efficient solution for axial piston pump fault diagnosis.
• This approach effectively handles non-linear and non-stationary vibration signals.
• The developed technique provides a valuable tool for predictive maintenance in hydraulic systems.