Application of Composite Spectrum in Agricultural Machines

Area of Science:

• Mechanical Engineering
• Signal Processing
• Agricultural Machinery

Background:

• Machine monitoring and fault detection are crucial for agricultural equipment efficiency.
• Vibration analysis is a common technique for identifying component health.
• Composite spectrum (CS) offers a data-fusion approach to simplify spectral analysis.

Purpose of the Study:

• To evaluate the effectiveness of composite spectrum (CS) techniques for monitoring and detecting rotating unbalances in combine harvester components.
• To assess the performance of CS with vibration data from non-optimal accelerometer placements.
• To compare different CS formulations (non-coherent, coherent, poly-coherent) for fault detection.

Main Methods:

• Acquired vibration signals from five combine harvester components under various working conditions (off, balanced, unbalanced) and speeds (idle, maximum RPM).
• Utilized four accelerometers placed in non-optimal locations on the combine harvester chassis.
• Analyzed four individual spectra and three types of CS (non-coherent, coherent, poly-coherent) from 18 different operational cases.

Main Results:

• Individual spectra successfully identified peaks corresponding to component rotational speeds.
• All three CS formulations retained essential information from individual spectra, reducing the overall data required for analysis.
• Coherent and poly-coherent CS demonstrated noise reduction compared to non-coherent CS and individual spectra.

Conclusions:

• Composite spectrum (CS) techniques simplify the monitoring and fault detection of rotating unbalances in agricultural machines.
• Effective detection is achievable with a reduced number of accelerometers, even in non-optimal positions.
• Coherent CS appears to be the most suitable formulation for monitoring and detecting rotating unbalances in agricultural machinery.