Analysis of methods of digital processing of vibration signals in diagnostics of tractor units

Abstract

This article provides an analytical review of digital vibration signal processing methods applicable to the diagnostics of engines, transmissions, power take-off shafts, bearing assemblies, chassis systems, and tractor frame and cab components. The article examines the sequence of measurement data conversion, from calibration and quality control of recordings to the generation of a diagnostic solution. It is shown that a single indicator of the overall vibration level does not provide reliable recognition of the technical condition under variable rotational speed, load, and road background. For identifying localized defects in rolling bearings, one of the most informative approaches is bandpass demodulation with envelope analysis. When diagnosing gears, envelope analysis is advisable to combine with analysis of the meshing frequency, its harmonics, and sidebands. During acceleration, braking, and significant changes in rotational speed, stationary spectral analysis should be supplemented with ordinal and/or time-frequency analysis. A functional-node classification of methods, a comparative matrix of their applicability, and a dual-loop processing scheme separating low-frequency assessment of overall tractor dynamics and high-frequency detection of local defects are proposed. Requirements for synchronizing signals with rotational speed, travel speed, transmission, load, and support surface type are formulated.