Multichannel Signals Reconstruction Based on Tunable Q-Factor Wavelet Transform-Morphological Component Analysis and...

Area of Science:

• Mechanical Engineering
• Signal Processing
• Data Science

Background:

• Rotating machinery condition monitoring faces challenges in high-speed remote transmission and large-capacity data storage.
• Accurate acquisition and reconstruction of vibration signals are crucial for effective fault diagnosis.

Purpose of the Study:

• To propose a novel multichannel signal reconstruction approach for rotating machines.
• To address limitations in data transmission and storage for condition monitoring.
• To improve the accuracy of fault characteristic detection in complex machinery.

Main Methods:

• Utilized tunable Q-factor wavelet transform-morphological component analysis (TQWT-MCA) to separate periodical impulses (LRC) from noise (HRC).
• Employed a sparse Bayesian iteration algorithm with a step-impulse dictionary for signal reconstruction.
• Applied compressed sensing (CS) principles for efficient data handling.

Main Results:

• The proposed TQWT-MCA and sparse Bayesian iteration method significantly improved reconstruction accuracy compared to existing methods.
• Achieved dramatic increases in data transmission and storage speeds.
• Successfully identified multiple fault characteristics in a gearbox, including bearing outer race, ball, and gear faults, with precise frequency identification.

Conclusions:

• The novel TQWT-MCA based compressed sensing approach effectively reconstructs multichannel signals for rotating machinery.
• This method enhances diagnostic capabilities by accurately identifying multiple fault types and frequencies.
• The approach offers a significant improvement in efficiency for data acquisition and processing in condition monitoring.